This article was downloaded by: [202.174.120.2] On: 24 November 2014, At: 07:45
Publisher: Institute for Operations Research and the Management Sciences (INFORMS)
INFORMS is located in Maryland, USA
Information Systems Research
Publication details, including instructions for authors and subscription information:
http://pubsonline.informs.org
An Empirical Analysis of the Contractual and Information
Structures of Business Process Outsourcing Relationships
Deepa Mani, Anitesh Barua, Andrew B. Whinston,
To cite this article:
Deepa Mani, Anitesh Barua, Andrew B. Whinston, (2012) An Empirical Analysis of the Contractual and Information
Structures of Business Process Outsourcing Relationships. Information Systems Research 23(3-part-1):618-634. http://
dx.doi.org/10.1287/isre.1110.0374
Full terms and conditions of use: http://pubsonline.informs.org/page/terms-and-conditions
This article may be used only for the purposes of research, teaching, and/or private study. Commercial use
or systematic downloading (by robots or other automatic processes) is prohibited without explicit Publisher
approval, unless otherwise noted. For more information, contact [email protected].
The Publisher does not warrant or guarantee the article’s accuracy, completeness, merchantability, fitness
for a particular purpose, or non-infringement. Descriptions of, or references to, products or publications, or
inclusion of an advertisement in this article, neither constitutes nor implies a guarantee, endorsement, or
support of claims made of that product, publication, or service.
Copyright © 2012, INFORMS
Please scroll down for article—it is on subsequent pages
INFORMS is the largest professional society in the world for professionals in the fields of operations research, management
science, and analytics.
For more information on INFORMS, its publications, membership, or meetings visit http://www.informs.org
Information Systems Research
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Vol. 23, No. 3, Part 1 of 2, September 2012, pp. 618–634
ISSN 1047-7047 (print) — ISSN 1526-5536 (online)
http://dx.doi.org/10.1287/isre.1110.0374
© 2012 INFORMS
An Empirical Analysis of the Contractual and
Information Structures of Business Process
Outsourcing Relationships
Deepa Mani
Indian School of Business, Hyderabad 500032, India, [email protected]
Anitesh Barua, Andrew B. Whinston
McCombs School of Business, The University of Texas at Austin, Austin, Texas 78712
{[email protected], [email protected]}
T
he emergence of information-intensive business process outsourcing (BPO) relationships calls for the study
of exchange performance beyond traditional considerations of the contractual structure that facilitates cooperative intent to include the information structure that facilitates the mutual exchange of information to enact
cooperative intent and coordinate actions between the user firm and the service provider. Yet, there has been
little analysis of the drivers and performance effects of the information structure of BPO relationships, including its linkages to the underlying contractual structure. This study integrates perspectives in neo-institutional
economics and information processing to develop and test the theoretical argument that the extent of use and
performance effects of the information structure of the BPO relationship are greater in time and materials BPO
contracts than in fixed-price BPO contracts. Survey data on 134 BPO relationships provide empirical support
for our hypotheses. The synergistic impact of incentives and information on BPO performance emphasizes that
their joint assessment is necessary to enhance the explanatory power of extant theories of organization. This
result also has implications for achieving maximum benefits from complex BPO arrangements that are more
likely to be characterized by time and material contracts.
Key words: BPO; outsourcing; governance; contractual structure; information structure; coordination;
information processing; performance
History: M. S. Krishnan, Senior Editor; T. S. Raghu, Associate Editor. This paper was received on January 27,
2007, and was with the authors 14 43 months for 3 revisions. Published online in Articles in Advance
November 11, 2011.
1.
Introduction
of value creation through BPO, evidenced by the large
number of firms that cite significant negative experiences with their outsourcing projects,2 and emergent
research (e.g., Rouse and Corbitt 2006) highlighting
the intrinsic complexity and high failure rate of BPO
initiatives. What, then, drives BPO performance?
The study of outsourcing performance has largely
been dominated by neo-institutional economic theories, which view efficacy of contract choice as central
to exchange performance. The outsourcing contract
aligns incentives between the user firm and the service provider to engender cooperative behavior that is
Business process outsourcing (BPO) refers to contracting with an external organization for the primary
responsibility of providing a business function (such
as accounting) or a certain business process within
a wider function (such as accounts payable within
accounting) (Linder et al. 2003). BPO is the fastest
growing segment of the outsourcing market,1 and represents the maturing of outsourcing from a cost saving tool for transaction-intensive business processes
to a flexible and powerful strategy for business transformation (Linder 2004, Gottfredson et al. 2005). Yet,
there are significant challenges inherent to the process
2
Seventy percent of the respondents in a 2005 survey by Deloitte
Consulting expressed significant dissatisfaction with their outsourcing projects. Similarly, a survey conducted by Bain Consulting
found that although 82% of large firms in North America engage
in BPO, almost half of the respondents say their outsourcing programs fall short of expectations. According to SAP INFO Solutions,
four out of five BPO contracts inked today will need to be renegotiated within two years. Furthermore, 20% of all such contracts
will collapse (Johnson 2006).
1
According to IDC, the worldwide BPO market is expected to grow
from $405 billion in 2003 to $682.5 billion in 2008. By 2008, the use
of external technology and business process services is likely to
move from a 2005 average of 12% of the corporation’s total costs
to 20% of total costs. Forecast growth rates for BPO are 10–15%
per annum (in contrast to 7.1% for technology infrastructure and
application outsourcing).
618
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
required to exert effort necessary to achieve outsourcing objectives and high levels of performance
(e.g., Lacity and Willcocks 1998). However, these
economic theories acknowledge that despite incentive alignment, ex-post adaptation failures might also
occur “because autonomous parties read and react
to signals differently, even though their purpose is to
achieve a timely and compatible combined response”
(Williamson 1991). This implies that outsourcing failures may stem not only from incentive conflicts but also
from cognitive conflicts or the firms’ inability to either
recognize and develop a shared understanding of
changes in the task environment or coordinate actions
to respond to such change (Gulati et al. 2005). This
class of adaptation problems, rooted in cognitive conflict between the participant firms, extends focus on
the contractual structure that facilitates mutual exchange
of rights and aligns incentives to include the information structure, or relational processes and technologies
that facilitate the mutual exchange of information to align
actions and achieve an integrated response, in order
to understand outsourcing performance.
The efficient exchange of information is especially
salient to BPO. In contrast to information technology outsourcing (ITO) that involves the management
of technological applications, infrastructure, and services, BPO involves the management of business processes that draw upon these technologies. Key inputs
and outputs of business processes are information,
and their successful execution often requires information coordination between process workers (Mani
et al. 2007). Furthermore, the dynamic relationship
between the business process and the business environment may result in greater rates of information
change in the task environment on an ongoing basis
than in ITO. This, in turn, results in larger amounts
of information that need to be communicated, processed, and translated into decisions in certain BPO
relationships.
Empirical research (Aron and Singh 2003, Mani
et al. 2007) in BPO suggests that the amount of information processed in the BPO relationship may be
represented along an increasing continuum with a
“data origin” and “knowledge end” (Aron and Singh
2003). These studies find that as the outsourced process moves along the continuum from being datato information- to knowledge-intensive, the nature of
information processing in the BPO relationship must
also make the allied shift from modular information
processing that occurs relatively independent of information changes in the value chain to significant, often
real-time, information processing that requires extensive information exchange between the user firm and
the service provider. However, the extent of information processing in the BPO relationship has been
largely analyzed by these studies independent of the
619
underlying contract, so that it is unclear whether and
to what extent the information structure impacts performance across different contractual archetypes. The
possibility that the BPO contract may impact coordination and information exchange between the user
firm and the provider is assumed in transaction cost
research (Williamson 1991), and alluded to in more
recent empirical research on outsourcing contracts
(e.g., Mayer and Argyres 2004). Yet, there is little theoretical or empirical research that analyzes how these
two dimensions of outsourcing relate to each other.
In this study, we integrate the premises of information processing theories and institutional economics
to advance and test the arguments that (a) the extent
of use of the information structure in the BPO relationship is related to the contractual allocation of risks
and incentives,3 and that (b) this alignment between
the contractual and information structures of the relationship impacts BPO performance.
The empirical testing of our hypotheses uses a twostage regression model that is estimated with survey data on 134 BPO relationships. BPO contracts are
estimated as one of two widely prevalent contracting forms—fixed price or time and materials contracts.
A significant portion of contractual risks of cost and
demand uncertainty in BPO are borne by the provider
(user firm) in fixed-price (time and materials) contracts. Thus, ceteris paribus, the risk-averse user firm
(provider) prefers a fixed-price (time and materials)
BPO contract. However, it is not contract choice, per
se, that is of interest to us in this study. We first
examine if the differences in risk and incentive allocation across fixed-price and time and materials contracts are associated with differences in the extent of
use of the information structure of the BPO relationship. We extend theories of information processing to
the interorganizational context to define the information structure in terms of joint action (Heide and John
1990, 1992; Bensaou and Venkatraman 1995), relational
emphasis on coordination versus control (Tushman and
Nadler 1978, Gulati and Singh 1998), and technological
capabilities (Argyres 1999, Brynjolffson 1994). Along an
increasing continuum, the values for these dimensions
represent a shift from an autonomous information
strategy, characterized by little information exchange,
to a tightly connected information strategy, marked
by the increased flow of information that enhances
the information capacity of the relationship. We then
assess if the performance impact of the three dimensions of information structure, as measured by service
satisfaction and perceived reduction in costs of process
ownership, varies across the two contract choices.
3
In this study, we do not posit or test the effect of the information
structure or its consideration on contract choice. See Poppo and
Zenger (2002) for analyses of the effect of relational processes on
contract choice in outsourcing.
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
620
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
The contributions of our study to theory are threefold. First, our findings emphasize that the joint
study of incentives and information is necessary to
enhance the explanatory power of extant theories of
organization—the contractual structure influences the
extent of investment in the information structure and
the effects of the latter on BPO performance. Second, the relation between the BPO contractual and
information structures points to an important function of the contract beyond the resolution of incentive
conflict. In addition to addressing exchange hazards
that stem from uncertainty perceived by the user firm
about its relationship with the provider, the BPO contract also facilitates the coordination of information
that addresses task uncertainty to integrate actions.
Finally, the study enumerates and validates some key
determinants of BPO performance. Although BPO is
increasingly viewed by modern organizations as critical to competitive success, there is little empirical
research that incorporates its distinctive nature and
form. Rouse and Corbitt (2006) find that during the
period January 1980 to June 2005, only 11 scholarly or
peer-reviewed articles reported empirical research on
BPO. They lament that “the absence of independent
empirical research means that decision makers choosing whether to outsource a business process have to
proceed on faith” (Rouse and Corbitt 2006, p. 586).
Our results address this gap in the extant literature
to improve theoretical understanding of the drivers of
BPO performance.
Our findings also contribute to outsourcing practice. A survey by the Outsourcing Center in 2005
found that while misaligned interests account for 15%
of outsourcing failures, poor governance accounts for
over 25% of these failures. Furthermore, a study by
Patni Computers found that these governance costs
are especially pronounced in the case of BPO. Our
study provides insights into what constitutes effective
governance beyond alignment of interests, including specific investments in information exchange that
help realize the benefits of BPO. We find that in contracts where there is significant information asymmetry on the true costs of executing the outsourced process, the user firm must commit to a tight relationship
with the provider and take the initiative to emphasize coordination and higher levels of information
exchange. This will help in effective task coordination, foster solidarity, and reduce information asymmetry to improve on contractual outcomes. Equally
important, the user firm must transfer ownership and
control of the business process to the provider when
the contract choice is consistent with its risk preferences. Our study finds that investments in information exchange in this context are redundant—while
having little marginal impact on satisfaction, they add
significantly to the costs of ownership of the outsourced process.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
2.
2.1.
Theory and Hypotheses
The Contractual Structure of the BPO
Relationship
Efficient contract design has long been considered
central to the success of outsourcing relationships
(e.g., Lacity and Willcocks 1998). The “right” contract
enables participant firms to overcome individual conflicts of interest and adapt effectively to disturbances.
Formulations of BPO contract problems involve variants of the two dominant contract types considered
in ITO (Banerjee and Duflo 2000, Gopal et al. 2003)—
fixed price contracts and time and materials contracts.
Fixed-price BPO contracts involve payment of a fixed
price per billing cycle or per unit of process output
(e.g., per claim processed or call fielded) per billing
cycle that is negotiated ex ante, while time and materials BPO contracts involve payment for the time and
effort expended by the provider in process delivery
during the billing cycle. Firms self-select one of these
contracts to minimize the economic tradeoff between
ex post inefficiencies of more incomplete time and
materials contracts and ex ante costs of design of more
complete fixed-price contracts, the relative magnitude
of which are systematically related to the nature of
the underlying task and relationship (Crocker and
Reynolds 1993).
The key ex post contractual inefficiencies in BPO
stem from assumptions of opportunistic behavior and
overruns in execution costs and demand that may be
attributable to the complexity and dynamism of the
outsourced process. The user firm and provider may
reduce these overruns through more precise specification and understanding respectively of process
requirements so that the process can be executed
with optimal resources at higher performance levels. However, the difficulty of third parties such as
courts in ascertaining the source of overruns (Banerjee
and Duflo 2000) and the discrepant shares of overrun borne by participant firms in the two contract
types on account of the underlying price mechanism
may engender unreliable behavior by the user firm
or provider. This, in turn, results in different allocation of risks to the firms across fixed-price and time
and materials contracts. In describing these contractual risks below, we adopt the user firm’s perspective
and assume that the participant decision makers are
risk averse.
The risk of cost overruns in BPO contracts mimics
that emphasized by prior research in ITO (Banerjee
and Duflo 2000, Gopal et al. 2003). These overruns are
an outcome of asymmetry in understanding between
firms of task concepts such as workflows, performance standards, or mutual interdependencies that
often results from the complexity of the outsourced
task and the ensuing inability to precisely specify
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
process requirements. They also result from variance
in process inputs and diversity of work processes
required to transform inputs to outputs that render it difficult to plan and allocate resources to process execution ex ante. This is because the output
of BPO relationships is information, which, unlike
products or applications, cannot be inventoried. Input
resources to service demand are not perfectly flexible, and thus the provider must choose a resource
allocation ex ante. However, demand for the outsourced process is only revealed ex post. In such
cases, lower-than-expected demand yields resource
wastage, while higher-than-expected demand necessitates resource acquisition. These resource costs
increase with the variety in resources required in
addressing unexpected demand—e.g., customer experience management for a financial services firm
would vary depending on the net worth of the customer; resources used to service the process for lowincome customers would not readily scale to meet
unexpected increase in high-net-worth consumers.
In time and materials BPO contracts, the user firm
bears the major share of the above cost overruns
resulting from process complexity and variety. Given
the asymmetry between participant firms on the true
costs of executing the outsourced process, the pricing mechanism in time and materials contracts entails
risks to the user firm of moral hazard and dissipation
of gains through costly bargaining or privately favorable redistribution of ex post surplus. Alternatively,
in fixed-price contracts, the payment to the provider
is not revised based on the latter’s cost experience,
and hence, the provider bears the risk of cost overruns. This is true irrespective of whether the output
is indivisible or variable—variable output transactions
in BPO (e.g., claims processed, calls fielded) are more
easily measurable and verifiable than input resources,
and fixed-price contracts compensate the provider for
the former at a fixed rate negotiated ex ante, while
time and materials contracts compensate the provider
for the latter. Thus, ex post efficiency losses to the user
firm from cost uncertainty and allied risks of moral
hazard and costly bargaining are lower in fixed price
BPO contracts than in time and materials contracts.
The above argument only considers the risk of cost
overruns that result from uncertainty in the average level of effort required to execute dynamic processes. However, BPO contracts also involve the risk
of demand overruns that result from uncertainty in
the volume of work to be executed for dynamic
processes. A potential risk of demand overruns in
fixed-price contracts is nonfulfillment of unexpected
demand that in turn, may impact customer satisfaction or operational efficiency. However, we assume
that nonfulfillment of unexpected demand for the
621
outsourced process is precluded by reputation concerns or continuity expectations in the BPO relationship. Moreover, when the volume of work cannot
be predicted accurately ex ante, the more restrictive
fixed-price contract where the price per transaction is
specified to be independent of future contingencies is
often expanded to include a floor and ceiling for the
volume of work to be performed.4 In the case of such
not-to-exceed demand fixed-price contracts, the risk
of nonfulfillment of unexpected demand is borne by
the user firm.
If contracting were costless, the user firm would
expend effort to anticipate contingencies in the task
environment that affect cost and demand overruns,
design responses to these contingencies, and precisely
implement these expectations in the contract. However, the cost to the user firm of describing ex ante
what it wants or the cost to the provider of understanding such expectations increases significantly in
complex, dynamic process environments, placing economic constraints on the ability of participant firms
to limit their exposure to risks of cost and demand
overruns through more complete fixed-price contracts
(Crocker and Reynolds 1993).
Furthermore, prior research in economics (Holmstrom and Milgrom 1991, Bajari and Tadelis 2001)
and IS (Aron and Singh 2005) has emphasized the
cost-quality tradeoff inherent to these two contracts—
output quality may be adversely affected by costsaving efforts of the provider. Thus, time and
materials contracts may be more appropriate in the
case of outsourced business processes, where output
process quality is a critical component of exchange
performance.
These risks, costs, and incentives inherent to the
two BPO contract types are outlined in Table 1. The
premise of this study is that in complex, dynamic BPO
relationships, where it is costly to implement more
complete fixed-price contracts or difficult-to-provide
incentives to the provider to constrain costs while not
compromising on output quality, the contract is limited in its ability to protect the user firm from cost
and demand risks. Banerjee and Duflo (2000, p. 993)
suggest that “when the first best cannot be achieved by
contractual means, it is possible to improve on the (contractual) outcome if the behavior of the firms and the clients
is at least partly norm-governed.” This study introduces
the information structure as an important dimension
of the BPO relationship that helps reduce information asymmetries between the firms to address ex post
contractual inefficiencies. We argue that the mutual
4
Knowledge@Wharton Business Process Outsourcing Survey:
Clients Want Strategic Partners That Help Cut Risk, Grow Customer Satisfaction.
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
622
Table 1
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
A Comparison of Fixed-Price and Time and Materials BPO Contracts
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Parameter
Fixed price
Time and materials
Contract price
Fixed price that is based on process output and
independent of input costs
Variable price that is dependent on input
resources or cost experience of the
provider
Contractual dimensions negotiated ex ante
• Price per billing cycle OR
• Price per transaction per billing cycle
Target resource expend per billing cycle
Contractual dimensions negotiated ex post
No ex post price revisions based on the cost
experience of the provider
Final resource expend per billing cycle
Compensation for cost overruns caused by lower
analyzability of the outsourced process
Provider not compensated for cost overruns
• Provider compensated for cost overruns
• Efficiency losses to the user firm due to
risks of:
Information asymmetry and moral
hazard
Costly bargaining and privately
favorable redistribution of surplus
Compensation for demand overruns caused by
higher variety of the outsourced process
Provider not compensated for effort to meet
demand overruns
• Provider compensated for effort to meet
demand overruns
• Efficiency losses to the user firm due to
risks of:
Information asymmetry and moral
hazard
Resource-consuming redistributive
tactics
Ex ante costs of contract design
Incentives for quality
Incentives for cost reduction
Higher
Lower
Higher
Lower
Higher
Lower
exchange of information through the information structure of the BPO relationship complements the limits
to adaptability of mutual exchange of rights through the
BPO contract to improve performance outcomes.
2.2.
The Information Structure of the BPO
Relationship
While the neo-institutional economics literature has
largely focused on analyzing the class of adaptation
problems that arise from bounded rationality and
allied uncertainty perceived by the user firm about its
relationship with the provider, information processing theorists (Galbraith 1973, Tushman and Nadler
1978) have contemporaneously engaged in studying
another class of adaptation problems that also arises
from bounded rationality but has its origin, not in
behavioral uncertainty, but rather in uncertainty in
organizational tasks.
The information processing view of the firm
(Galbraith 1973) offers that organizations must process information to cope with uncertainty in their
task environment, and that variation in organizational form represents variation in strategies to reduce
its information requirements, or increase its capacity to adapt to its inability to reduce information
requirements. Therefore, while the economic literature emphasizes the role of mutual exchange of rights
in the contract in effective adaptation, the information processing literature views the mutual exchange of
information as central to effective adaptation (Sobrero
and Schrader 1998). This is largely because of the
difference in the contextual emphasis of the two
theories—while the neo-institutional economics literature focuses on the organization of economic activity
across firm boundaries, the central ideas of the information processing view have largely been applied to
the study of organization design that best facilitates
intraorganizational adaptation.
Yet, this research paradigm is particularly salient
to the context of BPO. Efficient work design in
the BPO relationship requires that the user firm
and the provider overcome cognitive limitations and
develop a shared understanding of the outsourced
task and interdependencies required for process execution. Asymmetry in understanding task concepts
such as quality or performance standards needs to be
resolved so that the task can be executed with optimal resources and minimal overruns. While the BPO
contract aligns incentives to provide a framework for
cooperation, it does not ensure accurate interpretation or enactment of this framework to align actions
(Gulati et al. 2005). In an organization, culture and
affiliation enable members to develop shared meanings (Gioia and Thomas 1996) required for efficient
work design. However, since participant firms in the
BPO relationship are affiliated with different cultures,
information may fill the key role that culture serves
for organizations. The information structure helps in
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
the dissemination of process information across firm
boundaries so that an integrated pattern of behavior emerges to enact the contract at higher levels of
performance.
In the interorganizational context, the information structure additionally addresses uncertainty perceived by the user firm about its relationship with
the provider to reduce asymmetry on the true costs
of executing the outsourced process. Greater levels
of information exchange help participant firms gain
insights into the cost experience of executing the
outsourced process, including sources of cost and
demand overruns. This, in turn, helps the user firm
formalize controls and responses to these sources of
overruns to reduce bargaining costs and redistribution of surplus. In addition to reducing the potential
for opportunism, the information structure also helps
reduce the provider’s propensity for opportunism by
increasing the likelihood of tie formation and fostering mutual trust between participant firms.
Information processing theorists (Galbraith 1973,
McCann and Galbraith 1981) find that information
exchanges vary along three relational dimensions—
joint action, relational emphasis on coordination
(versus control), and technological capabilities. Joint
action enables information exchange to sequence,
schedule, and synchronize outsourced tasks. In doing
so, joint action increases the frequency of contact
between the user firm and provider and in turn, the
likelihood of tie formation between them. Joint effort
in key process areas also requires mutual adaptation and adjustment, engendering norms of solidarity
and flexibility (Poppo and Zenger 2002). Relational
emphasis on coordination reflects a lower degree of
formalization or control in the relationship (Bensaou
and Venkatraman 1995). This signals to the provider
that it is trusted or trustworthy to behave appropriately without such control (Ghoshal and Moran 1996,
Dyer and Singh 1998) thereby, promoting a cooperative environment and free exchange of information.
Lower levels of formalization also enable timely sharing of process information and clarification of task
outputs that must be integrated back into the user
firm’s value chain. While joint action and emphasis on coordination help participant firms develop a
shared understanding of changes in the information
environment of the outsourced process, technological
capabilities enable an accurate and timely incorporation of these changes in the execution of the process.
They establish a “standard grammar for communication”
(Argyres 1999) that increases the amount of information exchanged per unit of time between decision
makers to reduce information overload. This standard grammar also helps coordination to be carried
out without costly idiosyncratic investments, thereby
reducing the potential for provider opportunism
623
(Argyres 1999). Relational technologies enable coordination without extensive communication, thereby
reducing the costs of information exchange. Finally,
technological capabilities enable continuous measurement of process output and provider performance to
reduce monitoring costs and maximize surplus.
The above arguments emphasize that as the information needs of the BPO relationship increase, the
level of joint action, emphasis on coordination, and
technological capabilities also increases, and the information structure shifts from being mechanistic or
autonomous to being organic or highly connected
(Galbraith 1973, Tushman and Nadler 1978).
Research on contractual and information structures
has largely occurred independent of each other and
in different contexts. Recent research (e.g., Poppo and
Zenger 2002) points to possible complementarities
between contractual and relational design (that is a
dimension of the information structure). However,
these studies do not consider the role of a holistic information strategy, encompassing both relational
and technological design, or that this strategy might
vary systematically with the underlying contract type
to shape performance (Puranam and Gulati 2005).
This study addresses this gap in the literature to
address two questions: (a) First, does the extent of use
of the information structure vary with the information asymmetries and allied risks in the BPO contract?
(b) Second, are there significant performance gains
from such effort?
2.3.
The Relationship Between Contractual and
Information Structures
We argue that the allocation of risks and incentives
in time and materials contracts necessitates greater
information exchange between the user firm and the
provider, thereby providing important support for the
co-varying relationship between time and materials
contracts and user firm investments in each of the
three dimensions of the information structure—joint
action, relational emphasis on coordination, and technological capabilities.
First, in time and materials contracts, the user firm
is the residual claimant of ex post surplus. Hence, it
is less tolerant of information noise or inexact process information that increases cost overruns, and is
incentivized to invest in rich, customized information
exchanges that yield timely information sharing and
deep knowledge of the outsourced process. Such sharing of process knowledge is necessary on an ongoing
basis since the underlying process in time and materials contracts tends to be complex and dynamic and
hence, difficult to completely contractually describe
ahead of time. Each of the elements of the information
structure supports the user firm’s objective of accurate
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
624
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
and timely information sharing. Joint action is useful
for communicating informal and partly tacit process
knowledge to the provider; technological capabilities
improve the efficiency of such communication, enable
accurate incorporation of process knowledge in execution, and render the process more responsive to
changes in the business environment while lower
degrees of formalization provide the relationship with
the flexibility required for ongoing adaptation to such
changes.
Second, in time and materials contracts, the user
firm bears the risk of potential loss in efficiency
gains from costly bargaining and privately favorable redistribution of ex post surplus. The greater
the complexity of the outsourced process and the
more specialized the nature of its input resources,
the greater is the information asymmetry on the true
costs of executing the outsourced process and the
more pronounced is the risk of moral hazard and
efficiency losses. Furthermore, given exogenous process requirements, recommendations of the institutional economics literature to reduce the provider’s
chances for opportunism or the magnitude of ensuing
loss through reduction in the user firm’s stake in specific assets or diversification may be difficult to implement. An accepted feasible alternative is for the user
firm to attempt to reduce information asymmetries
and the provider’s propensity toward opportunism
by engendering trust through personal bonds, shared
norms, and values” (Nooteboom et al. 1997, p. 56). In
addition to serving as a conduit for transfer of process information, the information structure of the BPO
relationship supports this objective by promoting the
emergence of embeddedness in the relationship that
in turn, reduces the propensity for opportunism and
allied appropriation concerns.
Thus, greater levels of joint action, emphasis on
coordination and technological capabilities address
ex post inefficiencies in time and materials contracts
to improve performance. Prior research suggests that
“learning the information channels within a firm and
the codes for transmitting information through them”
involves costly investments in time and effort (Arrow
1974), and that the necessary magnitude of information exchange between the firms can help evaluate the
utility of such investments (Monteverde 1995). This
suggests that user firms will invest in using the information structure “only when significant hazards are
present” (Poppo and Zenger 2002). Absent these hazards, managerial time and effort are wasted in the
use of the information structure and engender performance expectations that are misplaced in the context
of simple, modular BPO relationships. Consequently,
we expect fixed price contracts that are more complete
and pose few risks to the user firm will be associated
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
with lower joint action, emphasis on coordination and
technological capabilities. We hypothesize:
Hypothesis 1A. The extent of use of joint action will
be greater in time and materials contracts than in fixedprice contracts.
Hypothesis 1B. The extent of relational emphasis on
coordination will be greater in time and materials contracts
than in fixed-price contracts.
Hypothesis 1C. The extent of use of technological
capabilities will be greater in time and materials contracts
than in fixed-price contracts.
2.4.
The Joint Impact of Contractual and
Information Structures on Performance
In this section, we examine the joint impact that the
contractual and information structures have on BPO
performance. In particular, we study contingent complementary and substitutional relationships between
the contractual and information structures of the BPO
relationship to theorize how choice of the former
could potentially impact the performance impact of
the latter.
The more complete incorporation of contingencies
that affect cost and demand overruns in fixed-price
contracts minimizes the need for ongoing control and
coordination (Joskow 1988, Heide 1994). To the extent
that the modes of control formalized in the fixed-price
BPO contract address contractual risks and preclude
opportunistic behavior, investments in coordination
structures, joint action, and technological capabilities
present additional costs with no additional benefits.
Furthermore, greater completeness of fixed-price contracts also suggests that the underlying outsourced
process is relatively simple with little information or
knowledge churn. Thus, ownership and control of the
process can be transferred to the provider and process output can be integrated back into the user firm’s
value chain with relative ease. In such cases, the contract, in addition to addressing relational uncertainty
and exchange hazards, also guides and legitimizes
information flows between the firms to addresses process uncertainty, thereby rendering additional channels for the mutual exchange of process information
costly and ineffective. Such redundancy of resources
renders management of the BPO relationship relatively inefficient.
Prior research (Ghoshal and Moran 1996, Sitkin and
Roth 1993) extends the above argument that more
complete contracts supplant the need to invest in
coordination processes and technologies to argue that
contractual completeness detracts from the effectiveness of such processes and technologies. A detailed,
precise fixed-price contract, while consistent with
the risk-averse user firm’s preferences, may signal
distrust that blunts the demands of solidarity and
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
mutuality of relational processes. This inconsistency
between the attributes of the contractual and information structures may adversely influence the performance effect of the latter.
While the information structure may substitute
for the fixed-price contract, it complements the limits to adaptability of time and materials contracts.
The information structure helps in the continual
transfer of critical process information and knowledge that enhance the provider’s understanding of
the outsourced process to reduce cost and demand
overruns. Furthermore, it fosters trust and information exchange that reduce information asymmetries
in the relationship and the provider’s propensity
for opportunism in the face of cost and demand
overruns. These behaviors reduce ex post inefficiencies but are not contractible. Thus, the information
structure reduces ex post contractual inefficiencies to
improve performance outcomes in time and materials
contracts.
The above theoretical arguments suggest that the
performance impact of the information structure of
the BPO relationship will be higher in time and materials contracts than in fixed-price contracts. Thus,
we posit:
Hypothesis 2A. The effect of joint action on BPO performance will be stronger in time and materials contracts
relative to fixed-price contracts.
Hypothesis 2B. The effect of relational emphasis on
coordination on BPO performance will be stronger in time
and materials contracts relative to fixed-price contracts.
Hypothesis 2C. The effect of technological capabilities
on BPO performance will be stronger in time and materials
contracts relative to fixed-price contracts.
3.
Empirical Analysis
3.1. Data Collection
The data for this study were obtained through a
survey of senior executives responsible for the management and review of outsourced business processes. Drawing on prior research (Halvey and Melby
2000, IDC report) and our discussions with outsourcing practitioners, we developed a list of ten business functions that are most commonly outsourced.
A structured questionnaire was developed based on
comprehensive reviews of the literature and initial
interviews with 20 BPO experts.5 These exploratory
interviews were conducted with the underlying
objective of assessing applicability of information
5
The subject experts comprised directors of strategic outsourcing
practices in Fortune 100 firms (in financial services, healthcare,
retail, and high-tech), outsourcing advisory consultants, leading
Indian offshore vendors, and academicians.
625
processing theories to inter-firm BPO relationships,
validating logical linkages between the constructs,
and obtaining more clarity of perspective on desirable sample characteristics. The interviews also influenced questionnaire design and component items,
especially those that were being adapted to the
BPO context. Consequent to these exploratory interviews, we developed a structured questionnaire that
was pre-tested with a total of 30 medium to large
organizations, market research firms, and academicians. A seven-point Likert scale was used for most
questions; however, some questions involved binary
choices.
The desired sample included small to large organizations across industries that had outsourced one
or more business processes. Our list of respondents
came from several active compilations of outsourcing
firms, industry association referrals, and outsourcing
advisory referrals. A technique deployed in related
research in surveying executives is “to define populations and response rates based on those who will
precommit to respond” (Poppo and Zenger 2002).
The normative response rates based on precommitted samples are as high as 40% (Anderson and Narus
1990, Poppo and Zenger 2002). Six-hundred precommitted surveys were mailed, with follow-up letters
five weeks later. We received a total of 145 valid
responses of which 134 were complete in all respects.
This response rate of approximately 24% was lower
than expected, and was likely due to the lengthy
and extensive nature of the questionnaire. However,
it is consistent with the rate found in other studies
(Mohr and Spekman 1994, Weiss and Anderson 1992).
The final sample was representative of a range of
outsourcing objectives for which there was sufficient
variance in exchange and task attributes.
All respondents were assured that their responses
would remain confidential and that results would
be reported only in aggregate, thereby addressing
privacy concerns and minimizing potential bias in
self-reported data. We also distributed the questions
measuring each construct across the survey to minimize response bias. To test for potential response
bias, we compared the firm size in the sample to
the larger population. The sample and population
did not appear to differ in terms of firm size. We
also compared questionnaires turned in early with
questionnaires turned in later along a number of
variables—firm size, industry, respondent position,
exchange attributes, and service satisfaction—to
check for nonresponse bias (Poppo and Zenger
2002, Armstrong and Overton 1977). This procedure assumes that late respondents share similar
attributes and biases with nonrespondents. No significant differences were found between early and
late respondents. We also checked for the presence
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
626
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
of common-method bias (Podsakoff et al. 2003) and
concluded that common-method bias did not likely
impact survey responses.
Finally, a section of the raw data was randomly
subject to cross validation. For a random sample of
25 user firms, we requested the respondent to identify the service provider for the purpose of a brief
interview. 10 firms obliged, and we interviewed the
providers to obtain relevant process information. The
two information sets in the dyad were mutually
consistent.
3.2.
Measures
3.2.1. Exchange Performance. The emergent and
confidential nature of objective measures of BPO performance rendered their use difficult in this study.
Consequently, we used multiple dimensions of service satisfaction as metrics of exchange performance.
Our use of service satisfaction is consistent with prior
measures of performance found in the organization
and strategy literature (Poppo and Zenger 1998, 2002;
Gulati et al. 2005). Satisfaction is an important issue in
working partnerships like BPO since it is a proxy for
perceived effectiveness (Poppo and Zenger 2002, Barber and Venkatraman 1986). Furthermore, given that
BPO is characterized by a range of business objectives
ranging from reduced costs to innovation to speed
to market, a single objective measure of performance
may be inadequate. Satisfaction is also a significant
determinant of future actions, including repeat business, positive word-of-mouth, and loyalty (Barber and
Venkatraman 1986). Thus, although not identical to
exchange performance, we expect that satisfaction of
the user firm will be strongly correlated with the
former.
We adapted established scales to measure service
satisfaction. We used a seven-point scale where “1”
represented “dissatisfied” and “7” represented “satisfied.” A total of 11 items were used to measure service
reliability, responsiveness, innovation, and systematization or the processes, procedures, systems, and
technology that make a service a seamless one.
We complement the use of service satisfaction with
that of perceived improvement in process efficiency.
In interviews with process owners, we found that process efficiency in BPO relationships are assessed using
a combination of metrics such as process costs, labor
productivity, process accuracy or error rate, and process turnaround time or cycle time. Given that nearly
all BPO contracts in our sample measured reduction
in process costs to assess process performance, we
included it as another measure of performance of the
BPO relationship. Process costs in the case of vertical
integration include activity costs of labor and overheads, material and handling costs, if any, cost of
invested capital, if any, and the opportunity cost of
outsourcing estimated by the strategic impact of BPO.
Process costs for BPO include one time and ongoing costs of BPO, and any anticipated future pricing
adjustments.
3.2.2. Explanatory Variables. In order to perform
a comparative analysis of performance across the
identified categories of BPO relationships, we define
contractual structure to be equal to one for cases
where a firm employs time and material contracts,
and equal to zero for cases where a firm uses fixedprice contracts. Thus, we use the following probit
model to estimate contract choice:
Pr4Yi = 15 = Pr4Yi > 05 = ê4‚0 Xi 50
(1)
Here, Yi represents the contract choice—time and
materials or fixed price—for transaction i. Xi is a vector of determinants of contract choice; ‚ is the vector of coefficients of these determinants; and ê4 · 5
is the standard normal cumulative distribution function. The determinants of contract choice that we
consider are process attributes of analyzability, variety, modularity, and information intensity, and relational
attributes of prior cooperative association, relative bargaining power of the user firm, contractual completeness, and contract length. We also control for size of
the user firm in our analyses. We argue that firms
self-select one of fixed-price or time and materials
contracts to minimize the tradeoff between ex post
contractual inefficiencies and ex ante costs of contract
design, the relative magnitude of which is systematically related to the abovementioned attributes of the
underlying task and relationship with the provider
(Crocker and Reynolds 1993). Variables that shift the
marginal costs of contracting upward result in less
complete time and materials contracts while variables that shift the marginal benefits of contracting
upward render more structured fixed price contracts
more attractive (Crocker and Reynolds 1993). Specific
relationships theorized between the process and relational attributes and contract choice are described in
Table 2.
Measures of the antecedents of contract choice and
dimensions of the information structure, including
joint action, relational emphasis on coordination, and
technological capabilities were all adapted from prior
literature. All measurement items were tested for
content validity through interviews and discussions
with outsourcing practitioners and service providers.
Reliability of all scales used were satisfactory with
Cronbach alpha values greater than 0.70.
3.2.3. Data Analysis. First, we use a structural
model to provide evidence of an integrative latent
representation of information structure as shown in
Figure 1. The information structure of the BPO relationship is conceptualized as a second-order construct
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
627
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
Table 2
Estimates for the Model of Information Structure of the BPO
Relationship
Parameter
estimate
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Construct
Information structure
Joint action
Relational emphasis on
coordination
Technological capabilities
Technological capabilities
Scope of IT use
Intensity of IT use
Sophistication of coordination
systems
Composite
reliability
Variance
extracted
0081
0065
00729∗∗∗
00780∗∗∗
Table 3
Probit Estimates for Choice of Contractual Structure
Time and Materials
Outsourced process characteristics
Process analyzability
Process variety
Process modularity
00781∗∗∗
0085
0062
00760∗∗∗
00771∗∗∗
00878∗∗∗
Information intensity
Relational characteristics
Prior cooperative association
Bargaining power of the user firm
comprising joint action, relational emphasis on coordination, and technological capabilities. Technological
capabilities is conceptualized as a first-order construct comprising scope of IT use, intensity of IT use,
and sophistication of coordination systems. Composite z-scores were created for all indicators to facilitate
combining and comparing different scales of measurement. Factor scores for the technological capabilities
and information structure constructs obtained at this
stage are used in subsequent analyses.
Subsequently, we analyze the determinants of contract choice, the influence of such choice on the
information structure and their joint influence on
BPO performance. Prior research (e.g., Hamilton and
Nickerson 2003) suggests that the inability to control for unobserved features of the exchange (such as
user firm or provider attributes, market conditions,
or embeddedness between the firms) that simultaneously influence the choice of contract as well as use
of the information structure and BPO performance
results in biased, inconsistent estimates. To account
for possibly endogenous choices of the contract, we
employ a two-stage regression model that is derived
from work in labor economics (Heckman 1979, Lee
et al. 1980, Smith 1980).
We first estimate a first-stage probit model that
represents the selection equation (Equation (1)). The
instrument variables used in the selection equation
Figure 1
Second-Order Construct of Information Structure of the BPO
Relationship
Joint action
Information
structure
Emphasis on
coordination
Contractual completeness
Continuity expectations
Information structure
Joint action
Relational emphasis on coordination
Technological capabilities
Firm size
Performance guarantees
Competitive intensity
Constant
Log likelihood
Pseudo R-square
∗
−00349∗
4001945
00272∗
4001615
−00010
4001665
−00159
4001475
−00590∗∗∗
4001695
−00528∗∗∗
4001445
−00452∗∗
4001855
00295∗
4001575
00124
4002255
00112
4002365
00165
4002205
−00318∗∗
4001275
−00480∗∗
4001725
−00326∗
4001695
00535∗∗∗
4001425
−50050
0043
p < 0010; ∗∗ p < 0005; ∗∗∗ p < 0001.
are performance guarantees offered by the service
provider and competitive intensity. The variables
used in the model, including the rationale for their
choice, are described in greater detail in Table 3.
Using the predicted probabilities from the probit
model, we construct the inverse Mills ratio, ‹ji , which
is included as a control variable in the second-stage
models of information structure and BPO performance. Second-stage models, which incorporate the
correction for self-selection, provide consistent and
unbiased estimates. All models report heteroscedasticity consistent robust standard errors that correct for
underestimation of variances resulting from potential
nonindependence among sample observations.
Scope of IT use
Technological
capabilities
Intensity of IT use
Coordination systems
4.
Results
Table 2 details the factor loadings for the second-order
information structure construct and the first-order
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
628
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
technological capabilities construct. All factor loadings are significant and of high magnitude. Furthermore, measures of composite construct reliability and
variance extracted for both constructs are greater than
the threshold levels of 0.70 and 0.50 respectively. The
goodness-of-fit measures also provide adequate support for the information structure model. The normed
chi-square is less than the threshold value of 3, the
GFI and NFI are greater than threshold values of 0.90
and the RMSEA is less than the threshold value of
0.05. We use the output matrix of factor score weights
to calculate a weighted average score for information
structure and technological capabilities that we use in
subsequent analyses.
Table 3 presents the results of the contract choice
model that are used to formulate the inverse Mills
ratio for the second-stage information processing and
performance models. Both instrumental variables,
performance guarantees and competitive intensity,
Table 4
Estimates for the overall Information Structure and Joint Action, Coordination, and Technological
Capabilities Across Contractual Structures
Information
structure
Contractual structure
Time and materials
Outsourced process characteristics
Process analyzability
Process variety
Process modularity
Information intensity
Relational characteristics
Prior cooperative association
Bargaining power of the user firm
Contractual completeness
Continuity expectations
Other controls
Firm size
Correction for self-selection (‹5
N
Adjusted R2
Model F
are significant predictors of contract choice. Furthermore, the results are consistent with the differences
in contractual risks between the two contracts. The
underlying process in fixed-price contracts is more
analyzable, less dynamic, and lends itself to a more
complete specification of tasks and responsibilities.
Greater completeness is also likely an outcome of
greater competence of the user firm in transacting
with the provider through prior association—the latter is positively associated with the choice of a fixedprice contract. Greater complexity and variety in task
execution and the lack of prior association that underlie the choice of more incomplete time and materials
contracts result in greater risks of cost and demand
overruns and ex post inefficiencies of costly bargaining and privately favorable redistribution of contractual surplus.
Table 4 presents the results of three treatmenteffect models (Shaver 1998) where the dependent
Joint
action
Emphasis on
coordination
00617∗∗∗
4001605
00555∗∗∗
00560∗∗∗
−00349∗∗∗
4000585
−00075
4000705
−00079
4000585
00177∗∗∗
4000595
−00269∗∗∗
−00453∗∗∗
−00179∗∗
−00031
−00144∗∗
00046
00433∗∗∗
4000685
00142∗∗
4000625
−00059
4000735
00074
4000585
−00025
4000535
−00434∗∗∗
4001405
134
0054
21053∗∗∗
Technological
capabilities
00508∗∗∗
−00150∗
00005
−00146∗∗
00152∗∗
00173∗∗∗
00133∗
00336∗∗∗
00406∗∗∗
00374∗∗∗
00183∗∗∗
00001
00210∗∗∗
−00216∗∗∗
00098
−00051
00023
−00076
00088
−00086
00050
−00042
−00398∗∗∗
−00401∗∗∗
134
0044
10090∗∗∗
134
0053
22022∗∗∗
−00365∗∗
134
0032
8069∗∗∗
Notes. Breusch-Pagan test of independence of equations: • 2 435 = 51095, p < 0001. Given the significant correlation
between the residuals of the three equations, we estimate the above system of equations using seemingly unrelated
regressions.
∗
p < 0010; ∗∗ p < 0005; ∗∗∗ p < 0001.
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
629
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
variables are the overall information structure of
the BPO relationship and its three dimensions—joint
action, coordination, and technological capabilities.
The simultaneous OLS estimation of the three dimensions through multivariate regression ignores potential correlation among the errors across equations;
however, because the three dimensions are correlated and the dependent variables are same across the
equations, there may be contemporaneous correlation
among errors across the three equations. Thus, we use
the seemingly unrelated regression (SUR) model to
estimate the three equations. Our use of SUR is consistent with results of the Breusch-Pagan test of independence of equations, which finds that correlation
among the residuals of the equations for joint action,
coordination, and technological capabilities is significant (p < 00001). The regression of overall information structure of the BPO relationship reports robust
standard errors. The adjusted R-square values for the
models range from 0.32 to 0.54.
The coefficient for self-selection is significant in all
models, indicating that the firms have self-selected
the pertinent contract and that unobserved heterogeneity that impacts self-selection also impacts information structure and its dimensions. The choice of
the time and materials contract is significant in all
four models, providing support for Hypotheses 1A–
1C: the extent of overall information exchanged, and
that facilitated through each of joint action, emphasis on coordination, and technological capabilities is
greater in time and materials contracts than in fixedprice contracts. The finding that differences in the
three dimensions of information structure across the
two contracts are significantly different from zero is
robust to alternative estimation methods, including a
joint test of the effect of contract choice across the
three models through standard multivariate regression and OLS estimation with robust standard errors.
The results for the control variables are consistent
with prior research. Lower process analyzability and
prior cooperative association positively impact the
levels of joint action, coordination, and technological investments. Modularity of the outsourced process
suggests that it can be disaggregated from the value
chain and managed and executed independently by
the service provider with relatively few inputs from
the user firm. This explains its negative impact on
joint action. It has been shown in prior research (e.g.,
Bakos and Brynjolfsson 1993) that when the user firm
has no bargaining power, it does not get an adequate
share of the noncontractible surplus; yet if it has all
the bargaining power, then service providers might
not make any noncontractible investments, engendering little surplus to bargain over. These studies find
that in order to decrease the opportunism risks that
stem from the provider’s concerns that returns to its
investment will be expropriated in ex post bargaining, the user firm invests in increasing the supplier’s
bargaining power through development of close relations and this approach to partnering is especially
facilitated by the use of technology in the relationship
(Clemons et al. 1993). The positive impact of the user
firm’s bargaining power on the level of joint action
and technological investments is consistent with this
theoretical argument.
Table 5 reports results for the performance effect
of the three dimensions of information structure.
Model I provides results for service satisfaction while
Model II provides results for perceived reduction in
costs of process ownership. The correlation between
service satisfaction and perceived reduction in process
costs is 0.61 (p < 0001), emphasizing that satisfaction
is a reasonable proxy for the successful realization of
efficiency outcomes. F -tests for all models reject the
hypothesis that the predictors are jointly insignificant
Table 5
Performance Effect of Joint Action, Coordination, and
Technological Capabilities Across Contractual Structures
Service satisfaction
Process ownership
costs
Information
structure
Fixed
price
Fixed
price
Joint action
00208
4001545
−00069
4001565
00035
4001265
00354∗∗
4001385
00265∗
4001415
00200
4001405
−00114
4001875
−00293
4002365
−00063
4001455
00400∗∗∗
4001395
−00032
4001765
00304∗∗
4001515
−00012
4001295
00054
4001235
00254∗
4001355
00386∗∗∗
4001175
00064
4001045
00015
4001035
−00158
4001605
−00129
4001775
00101
4001395
00100
4001485
00144
4001195
00183
4001105
−00174
4001235
00055
4001085
−00045
4001685
00081
4001125
00385
4002365
−00142
4001485
00188
4001135
−00108
4001095
Relational emphasis
on coordination
Technological
capabilities
Process analyzability
Process variety
Process modularity
Information intensity
T&M
Prior cooperative
association
Bargaining power of
the user firm
00330∗∗
4001365
00340∗∗∗
4001225
00277∗∗
4001085
00020
4000885
Contractual
completeness
Continuity
expectations
00173
4001525
00316∗∗∗
4001115
00076
4001015
00169∗
4000965
00590∗∗∗
4001935
00265∗
4001535
T&M
−00031
4001165
−00040
4000895
Firm size
−00279∗∗
4001275
00176∗
4000985
−00035
4001965
00186∗∗
4000875
Correction for
self-selection 4‹5
N
Adjusted R2
Model F
−00075
4002095
51
0045
4037∗∗∗
−00074
4002225
79
0062
11047∗∗∗
−00324
4002665
48
0022
2009∗∗
00075
4002435
79
0042
5064∗∗∗
∗
p < 0010; ∗∗ p < 0005; ∗∗∗ p < 0001.
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
630
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
(p < 0001). Potential multicollinearity problems were
investigated by examining variance inflation factors
(VIFs) and condition indices for the predictor variables. Near multicollinearity is a problem if VIF > 10
or the maximum condition index (CI) for the model
>30. An analysis of these measures confirmed that
the three models were not biased by multicollinearity.
The adjusted R-Square values for the models range
from 0.22 to 0.62.
We test Hypotheses 2A–2C by comparing the significance of coefficients of the three dimensions of information structure across the two contract choices in
Table 5. We find that the impact of joint action and
coordination on service satisfaction and the impact of
joint action and technological capabilities on reduction in process costs are stronger in time and material contracts than in fixed-price contracts. Thus, our
results for the model of satisfaction support Hypotheses 2A and 2B while our results for the model of
reduction in process costs support Hypotheses 2A
and 2C.
Table 6 reports the results of analysis of the impact
of the overall information structure of the BPO relaTable 6
Performance Effect of Overall Information Structure Across
Contractual Structures
Service satisfaction
Process ownership
costs
Fixed
price
T&M
Fixed
price
Information
structure
00102
4001485
00670∗∗∗
4001095
−00363∗
4002025
00574∗∗∗
4001375
Process
analyzability
Process variety
00023
4001265
00037
4001225
00369∗∗∗
4001155
00036
4001015
−00125
4001455
−00159
4001655
00139
4001405
00110
4001235
00226∗
4001325
−00183
4001205
00309∗∗
4001405
00321∗∗
4001205
00097
4001345
−00000
4001035
00056
4001065
00298∗∗∗
4001105
00041
4000965
00054
4001055
00087
4001295
−00053
4001635
00345
4002155
−00143
4001445
00543∗∗∗
4001855
00149
4001135
00065
4001165
00205∗
4001195
−00012
4001055
−00071
4001085
00156
4000985
00250∗
4001415
−00041
4000855
Process modularity
Information intensity
Prior cooperative
association
Bargaining power of
the user firm
Contractual
completeness
Continuity
expectations
00333∗∗∗
4001075
T&M
Firm size
−00322∗∗
4001275
00185∗
4001005
−00035
4001835
00180∗∗
4000875
Correction for
self-selection 4‹5
N
Adjusted R2
Model F
−00066
4001945
51
0045
9010∗∗∗
−00107
4002265
79
0061
17085∗∗∗
−00306
4002535
48
0024
2079∗∗∗
00054
4002605
79
0041
7072∗∗∗
∗
p < 0010; ∗∗ p < 0005; ∗∗∗ p < 0001.
tionship on satisfaction and reduction in process
costs across the two contracts. A comparison of the
coefficient of information structure across both contract choices confirms that the effect of the overall
information structure on satisfaction and reduction in
process costs is significantly greater (p < 0001) in time
and materials contracts than in fixed-price contracts.
Furthermore, the impact of the information structure
on reduction in process costs is negative and significant in fixed-price contracts. This suggests that investments in information resources might be costly and
redundant in such cases, and might detract managerial attention, thereby adversely impacting the costs
of ownership of the outsourced process.
5.
Discussion and Conclusion
The outsourcing of value chain functions has gained
significant momentum as organizations increasingly
outsource broader business processes—particularly
information-intensive ones such as human resources,
finance and accounting, supply chain management,
and customer care—to achieve diverse strategic
objectives. BPO necessitates expanding the view of
outsourcing beyond a contractual transaction that
reduces production costs to include an information
network that enhances the user firm’s capacity to process information. While each of the contractual and
information structures has been investigated separately under different research paradigms and business contexts, our objective in this study has been to
investigate the relationship between these distinctive
dimensions in the context of BPO relationships.
Our empirical analysis confirms that the extent
of use and the performance impact of information processing mechanisms are stronger in time
and materials contracts than in fixed-price contracts.
Our results for the relationship between the contract
and information structure address the broader issue
of providing incentives for noncontractible investments. In particular, our analysis tests whether the
relatively complete nature of fixed-price contracts
that renders the provider the residual claimant of
ex post surplus provides a superior framework for
noncontractible investments in information processing (Hypotheses 1A–1C), and whether the incentives inherent to fixed-price contracts complement the
emergent information structure to enhance exchange
performance (Hypotheses 2A–2C). We find that the
impact of joint action and coordination on satisfaction
and that of joint action and technological capabilities
on perceived reduction in process costs are significant
in time and materials contracts. Thus, the division
of risks and incentives inherent to time and materials contracts complement emergence of the information structure and its subsequent impact on BPO
performance.
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
Our results hold implications for theory development. They emphasize that problems of governance
and information arise simultaneously in modern
outsourcing relationships and represent interlinked
explanations of outsourcing performance. Thus, moving forward, as firms and business functions become
increasingly information intensive,6 the joint assessment of contractual and information structures is necessary to yield a more holistic picture of the effects
of boundary decisions, and enhance the explanatory
power of extant theories of organization.
We find that although the information structure in
fixed-price contracts does not impact satisfaction, it
negatively influences process costs in these contracts.
When the user firm invests in high levels of joint
action, coordination and technological capabilities
in the BPO relationship, managerial attention and
involvement, a necessary input to this relational
dimension but often a constrained organizational
resource, likely becomes overloaded and inadequate
to sustain joint action at high levels of performance.
This is exacerbated by the fact that the user firm is not
the residual claimant of ex post surplus in fixed-price
contracts, so that the transaction and bureaucratic
costs associated with increased levels of joint action
outweigh the user firm’s marginal benefits, thereby
reducing the performance of the BPO relationship.
Our results for the interactions between the contractual and information structure of the exchange
also extend transaction cost analyses beyond traditional considerations of incentive conflict. The results
suggest that contracts differ not only in their ability to align incentives between the user firm and the
provider but also in their ability to motivate and coordinate information processing activities across firm
boundaries in an accurate and timely way. For example, a law firm in our sample outsourced training and
development to a leading provider of managed learning services and hosted learning and performance
technology. A key responsibility of the provider
was workplace compliance training programs for the
law firm’s employees and customers. The provider’s
extant customers accessed a Web-enabled training
module for compliance training. However, given that
employment law was one of the key practices of
the law firm and that the law firm’s employees had
unique training needs in employment, human relations, and benefits matters, the extant compliance
training modules of the provider would not suffice.
In this case, the technology firm could not afford to
6
Sinha and Van de Ven (2005) note that knowledge-intensive services are the dominant form of work in the industrialized world.
Similarly, Quinn et al. (1997) report that three-fourths of all economic activity involves the management of intellectual activities
and their interface to service outputs.
631
assume sole ownership of the process, and hence,
entered into a time and materials contract that shifted
a portion of the risk to the law firm. The latter
worked in close cooperation with the provider to
develop a learning content management system that
incorporated personnel handbooks, covenants not to
compete, and confidentiality agreements of the firm
and was customized to meet industry-specific needs
of the law firm’s customers. Thus, the contract not
only addressed exchange hazards, but also provided a
cooperative framework for superior coordination and
information processing.
This study is subject to certain limitations. Our
data is cross-sectional in nature, and our measures of
exchange performance are based on perceptual items
that do not directly account for the economic costs
and benefits of BPO. However, given the nascent and
topical nature of BPO, such data are difficult to obtain
from user firms for reasons of confidentiality. Furthermore, given the heterogeneity in BPO objectives (say
innovation versus increased revenue), allied strategic
importance of the outsourced processes, and ensuing
timeline of returns from outsourcing, it is difficult to
reconcile the BPO relationships in terms of a singular
measure of economic performance. In this regard, our
measures of service satisfaction and perceived reduction in process costs create a defined anchor point
and represent the best alternative. Measures of our
controls too are self-reported and more susceptible
to perceptual biases and distortions. However, given
the difficulty in obtaining observed measures for variables such as the number of alternative providers
(competitive intensity), perceptual measures provide
the best alternative.
The cross-sectional nature of our data also limits our ability to capture important learning effects
that develop in the BPO relationship over time. Prior
research in strategy (Anand and Khanna 2000) finds
that with experience accumulation, firms learn to better manage complexity in alliances. This is likely true
of BPO relationships as well. Future research could
examine the influence of learning effects on choice of
information structure and performance.
Similarly, ex ante considerations of information
structure may also influence the choice of outsourcing contract. Although this argument finds limited
support in our empirical analysis, the cross-sectional
nature of our data limits our ability to test this argument comprehensively. Future research could focus
on and test the complex interrelationships between
these constructs.
Despite these limitations, we believe that this study
makes important contributions to the literature on
organization of outsourcing relationships. Data on
transaction and relational characteristics of a range
of BPO relationships allows us to take an early step
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
632
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
towards understanding what drives the performance
of this rapidly growing outsourcing form. In doing so,
we complement and extend the rich literature on firm
boundaries. Consonant with research that emphasizes the increasingly information-intensive nature
of business processes, our results show that the
contractual structure, the information structure, and
their interaction influence BPO performance; the contract emerges, not just as a governance mechanism that addresses incentive conflict, but also as
an important mode of organization of information
that addresses cognitive conflict. Applied to an organizational context, the framework yields an important understanding of the value of BPO and enables
firms to define and enact a holistic governance strategy that positively influences firm profitability and
competitiveness.
Appendix
Firms self-select one of fixed-price or time and materials
contracts to minimize the tradeoff between ex post contractual inefficiencies and ex ante costs of contract design, the
relative magnitudes of which are systematically related to
the nature of the underlying task and the characteristics
of the relationship between the user firm and the provider
(Crocker and Reynolds 1993).
The process characteristics considered in this study are
analyzability, variety, modularity, and information intensity.
The lower the analyzability of the outsourced process, the
more difficult it is for the user firm to contractually specify
and communicate process requirements to the provider and
the more difficult it is for the provider to understand these
requirements. Similarly, the lower the modularity of the outsourced process, the more difficult it is to disaggregate the
business process from the user firm’s value chain and transfer ownership and control to the provider through a more
complete fixed-price contract. The greater the variety in the
inputs to the outsourced process and the means to convert
these inputs to outputs, the greater are the costs of designing more complete fixed-price contracts—variety necessitates systematic allocation of resources to identify diverse
action-outcome contingencies, and incorporate paths that
are best aligned with the objectives of the outsourcing initiative. Similarly, the greater the information churn in the business process, the more hypothetical are the contingencies in
the task environment since they depend on the nature and
content of new information that may emerge in response
to changes in the business environment. In such cases, it is
costly, even difficult to contractually specify a precise division of labor and responsibilities. Thus, we expect that the
greater the variety and information intensity of the outsourced process and the lower the analyzability and modularity of the outsourced process, the greater is the likelihood
of choice of a time and materials contract.
The relational variables considered in our analysis
include prior cooperative association, relative bargaining power
of the user firm, contractual completeness, and contract length.
On the one hand, Crocker and Reynolds (1993) find that
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
prior association reduces the likelihood of provider opportunism, encouraging the adoption of less complete contracts. On the other hand, firms, through prior cooperative
association, may have developed channels and communication codes for exchange of information. Greater competence
in transacting with each other makes it easier for the firms
to contractually specify the management of the interface
between them through more complete contracts. Similarly,
on one hand, the complete specification of contingencies
and time of performance in longer term contracts may be
difficult and costly. On the other hands, expectations of continuity of the relationship inherent to longer term contracts
may weaken the association between uncertainty and contractual incompleteness, engendering more completeness in
long-term contracts. We test these conflicting predictions
by examining the influence of prior cooperative association
and expected contract length on contract choice. Gopal et al.
(2003) find that user firms (providers) with considerable
bargaining power would be able to negotiate a fixed-price
(time and materials) contract independent of task attributes
that is aligned with their risk preferences. Thus, we test the
impact of the relative bargaining power of the user firm on
the choice of outsourcing contract. Finally, we test whether
fixed-price contracts indeed involve a more complete specification of obligations and responsibilities.
We also control for size of the user firm. Larger firms
have superior financial, technological, and human resource
endowments that render the design of complex contracts
more efficient and less expensive. Thus, we expect that
larger firms will design more complete contracts. The
instrument variables used in our model of contract choice
include performance guarantees and competitive intensity.
Performance guarantees assure delivery of specific performance outcomes. The ensuing completeness is more representative of a fixed-price contract. The rationale behind
the use of competitive intensity as an instrument is that
increased levels of competition in service provision provide
the user firm with diverse alternatives, thereby reducing the
provider’s bargaining power and increasing the probability
of choice of a fixed-price contract (Gopal et al. 2003).
References
Anand, B., T. Khanna. 2000. Do firms learn to create value? The
case of alliances. Strategic Management J. 21 295–315.
Anderson, J. C., J. A. Narus. 1990. A model of the distributor’s firm
and manufacturer firm working partnerships. J. Marketing 54
42–58.
Argyres, N. 1999. The impact of information technology on coordination: Evidence from the B-2 “stealth” bomber. Organ. Sci.
10(2) 162–180.
Armstrong, J. S., T. Overton. 1977. Estimating nonresponse bias in
mail surveys. J. Marketing Res. 14 396–402.
Aron, R., J. Singh. 2003. IT enabled strategic outsourcing: Knowledge intensive firms, information work and the extended organizational form. Knowledge@Wharton, The Wharton School,
University of Pennsylvania, Philadelphia.
Aron, R., J. V. Singh. 2005. Getting offshoring right. Harvard Bus.
Rev. 8(12) 135–143.
Arrow, K. J. 1974. The Limits of Organization. W. W. Norton & Co.,
New York.
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
Bajari, P., S. Tadelis. 2001. Incentives versus transaction costs:
A theory of procurement contracts. RAND J. Econom. 32(3)
387–407.
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
Bakos, J. Y., E. Brynjolfsson. 1993. From vendors to partners:
The role of information technology and incomplete contracts in buyer-supplier relationships. J. Organ. Comput. 3(3)
301–328.
Banerjee, A. V., E. Duflo. 2000. Reputation effects and the limits
of contracting: A study of the Indian software industry. Quart.
J. Econom. 115(3) 989–1017.
Barber, M. B., M. Venkatraman. 1986. The determinants of satisfaction for a high envelopment product: Three rival hypotheses
and their implications in the health care context. Adv. Consumer
Res. 13 317–320.
Bensaou, M., N. Venkatraman. 1995. Configurations of interorganizational relationships: A comparison between U.S. and
Japanese automakers. Management Sci. 41(9) 1471–1492.
Brynjolffson, E. 1994. Information assets, technology and organization. Management Sci. 40 1645–1663.
Clemons, E. K., M. C. Row. 1993. Limits to inter-firm coordination
through information technology: Results of a field study in
consumer packaged goods distribution. J. Management Inform.
Systems 10(1) 73–95.
Crocker, K., K. J. Reynolds. 1993. The efficiency of incomplete contracts: An empirical analysis of Air Force engine procurement.
RAND J. Econom. 24(1) 126–146.
Dyer, J. H., H. Singh. 1998. The relational view: Cooperative strategy and sources of interorganizational competitive advantage.
Acad. Management Rev. 23(4) 660–679.
Galbraith, J. R. 1973. Designing Complex Organizations. AddisonWesley, Reading, MA.
Ghoshal, S., P. Moran. 1996. Bad for practice: A critique of the transaction cost theory. Acad. Management Rev. 21(1) 31–47.
Gioia, D. A., J. B. Thomas. 1996. Identity, image and issue interpretation: Sensemaking during strategic change in academia.
Admin. Sci. Quart. 41 370–403.
Gopal, A., K. Sivaramakrishnan, M. S. Krishnan, T. Mukhopadhyay.
2003. Contracts in offshore software development: An empirical analysis. Management Sci. 49(12) 1671–1683.
Gottfredson, M., R. Puryear, S. Phillips. 2005. Strategic sourcing
from periphery to the core. Harvard Bus. Rev. 8(2) 2–9.
Gulati, R., J. A. Nickerson. 2008. Interorganizational trust, governance choice, and exchange performance. Organ. Sci. 19(5)
688–708.
Gulati, R., H. Singh. 1998. The architecture of cooperation: Managing coordination costs and appropriation concerns in strategic
alliances. Admin. Sci. Quart. 43 781–814.
Gulati, R., P. R. Lawrence, P. Puranam. 2005. Adaptation in vertical
relationships: Beyond incentive conflict. Strategic Management J.
26 415–440.
Halvey, J. K., B. M. Melby. 2000. Business Process Outsourcing—
Process, Strategies and Contracts. John Wiley & Sons, New York.
Hamilton, B. A., J. A. Nickerson. 2003. Correcting for endogeneity in strategic management research. Strategic Organ. 1(1)
51–78.
Heckman, J. 1979. Sample selection bias as a specification error.
Econometrica 47(1) 153–161.
Heide, J. B. 1994. Inter-organizational governance in marketing
channels. J. Marketing 58(1) 71–85.
Heide, J. B., G. John. 1990. Alliances in industrial purchasing: The
determinants of joint action in buyer-supplier relationships.
J. Marketing Res. 27(1) 24–36.
Heide, J. B., G. John. 1992. Do norms matter in marketing relationships? J. Marketing 56(2) 32–44.
633
Holmstrom, B., P. Milgrom. 1991. Multi-task principal-agent problems: Incentive contracts, asset ownership and job design.
J. Law, Econom. Organ. 7(special issue) 24–52.
Jensen, M., W. Meckling. 1976. Theory of the firm: Managerial
behavior, agency costs and ownership structure. J. Financial
Econom. 3 305–360.
Johnson, L. K. 2006. Successful business process outsourcing (Intelligence Section). Sloan Management Rev. 47(2) 5–6.
Joskow, P. L. 1988. Asset specificity and the structure of vertical
relationships: Empirical test of transaction cost analysis. J. Law,
Econom. Organ. 4 121–139.
Lacity, M. C., L. P. Willcocks. 1998. An empirical investigation of
information technology sourcing practices: Lessons from experience. MIS Quart. 22 363–408.
Lee, L., G. Maddala, R. Trost. 1980. Asymptotic covariance matrices of two-stage probit and two-stage tobit methods for
simultaneous equation models with selectivity. Econometrica 48
491–503.
Linder, J. C. 2004. Transformational outsourcing. Sloan Management
Rev. 45(2) 52–58.
Mani, D., A. Barua, A. Whinston. 2007. An empirical analysis of the
impact of information capabilities design on business process
outsourcing performance. Working paper, University of Texas
at Austin, Austin.
March, J. G., H. A. Simon. 1958. Organizations. Wiley, New York.
Mayer, K. J., N. S. Argyres. 2004. Learning to contract: Evidence from the personal computer industry. Organ. Sci. 14(4)
394–410.
McCann, J., J. R. Galbraith. 1981. Interdepartmental relations. P. C.
Nystrom, W. H. Starbuck, eds. Handbook of Organizational
Design, Vol. 1. Oxford University Press, New York, 60–84.
Mohr, J., R. Spekman. 1994. Characteristics of partnership success:
partnership attributes, communication behavior, and conflict
resolution techniques. Strategic Management J. 15(2) 135–142.
Monteverde, K. 1995. Technical dialog as an incentive for vertical integration in the semiconductor industry. Management Sci.
41(10) 1624–1639.
Nooteboom, B., J. Berger, N. G. Noorderhaven. 1997. Effects of trust
and governance on relational risk. Acad. Management J. 40(2)
308–338.
Podsakoff, P. M., S. B. MacKenzie, J.-Y. Lee, N. P. Podsakoff.
2003. Common method biases in behavioral research: A critical
review of the literature and recommended remedies. J. Appl.
Psych. 88(5) 879–903.
Poppo, L., T. Zenger. 1998. Testing alternative theories of the
firm: Transaction cost, knowledge-based, and measurement
explanations for make-or-buy decisions in information services. Strategic Management J. 19(9) 853–877.
Poppo, L., T. Zenger. 2002. Do formal contracts and relational governance function as substitutes or complements? Strategic Management J. 23 707–725.
Puranam, P., R. Gulati. 2005. The social and contractual structure of
vertical relationships. SSRN Working paper, London Business
School, London.
Quinn, J. B., J. J. Baruch, K. A. Zien. 1997. Innovation Explosion:
Using Intellect and Software to Revolutionize Growth Strategies.
Free Press, New York.
Rouse, A. C., B. J. Corbitt. 2006. Business process outsourcing: The
hysteresis effect and other lessons. R. Hirschheim, A. Heinzl,
J. Dibbern, eds. Information Systems Outsourcing: Enduring
Themes. New Perspectives and Global Challenges. Springer, Berlin,
583–602.
Shaver, J. M. 1998. Accounting for endogeneity when assessing
strategy performance: Does entry mode choice affect FDI survival? Management Sci. 44(4) 571–585.
Downloaded from informs.org by [202.174.120.2] on 24 November 2014, at 07:45 . For personal use only, all rights reserved.
634
Mani et al.: An Empirical Analysis of the Contractual and Information Structures of BPO Relationships
Sinha, K. K., A. Van de Ven. 2005. Designing work within and
between organizations. Organ. Sci. 16(4) 389–408.
Sitkin, S. B., N. L. Roth. 1993. Explaining the limited effectiveness
of legalistic “remedies” for trust/distrust. Organ. Sci.: J. Inst.
Management Sci. 4(3) 367–392.
Smith, J. P. 1980. Introduction. J. P. Smith, ed. Female Labor Supply:
Theory and Estimation. Princeton University Press, Princeton,
NJ, 3–23.
Sobrero, M., S. Schrader. 1998. Structuring inter-firm relationships:
A meta-analytic approach. Organ. Stud. 19 585–615.
Information Systems Research 23(3, Part 1 of 2), pp. 618–634, © 2012 INFORMS
Tushman, M. L., D. A. Nadler. 1978. Information processing as an
integrating concept in organizational design. Acad. Management
Rev. 3(3) 613–624.
Weiss, A. M., E. Anderson. 1992. Converting from independent to
employee salesforces: The role of perceived switching costs.
J. Marketing Res. 29(1) 101–115.
Williamson, O. E. 1991. Comparative economic organization: The
analysis of discrete structural alternatives. Admin. Sci. Quart.
36 269–296.