Sunday, February 19, 2012

Outsourcing decision support: a survey of

Outsourcing decision support: a survey of
benefits, risks, and decision factors
and decision factors
Tibor Kremic
NASA Glenn Research Center, Cleveland, Ohio, USA, and
Oya Icmeli Tukel and Walter O. Rom
Operations Management Department, College of Business Administration, Cleveland State University, Cleveland, Ohio, USA
Purpose – The purpose of this study is twofold. The first is to provide a structured review of the vast amount of outsourcing literature that has
accumulated in the past two decades using a decision support framework. The second purpose is to statistically analyze the contents of the studies to
identify commonalities as well as gaps, in order to suggest directions for future research.
Design/methodology/approach – The contents of more than 200 publications are analyzed using a variety of approaches. A decision support
framework is used to first classify whether the studies address outsourcing benefits, risks, motivations or factors. Next, each classification is further
described by the type of benefits, risks, etc. Additional relevant contents such as type of organization, and the location of the outsourcing practice are
also considered. Multivariate analyses consisting of cross tabulations, chi-square testing and cluster analysis are used for categorizing the studies with
the aim of identifying relationships among the studies which are not apparent when they are considered individually.
Findings – A number of trends and relationships are identified. For example, most studies focus on US for-profit organizations and are typically
theoretical, discussing benefits, risks and motivators. On the other hand, the research on outsourcing practices of non-profit organizations, where
objectives for outsourcing are typically politically driven, is found to be scarce. Furthermore, the results of the cluster analysis indicate that the studies
can be grouped into six clusters where the five small clusters are characterized by strong relationships with a few variables while the large cluster is
characterized by variables that are not addressed in the studies.
Practical implications – Outsourcing has become commonplace in today’s businesses. In addition to outsourcing in profit seeking organizations,
there is considerable outsourcing effort in governmental and non-profit organizations also. It is not easy for managers who are exploring outsourcing
opportunities for the very first time and academicians who want to build upon existing studies to search the literature to find what they are looking for.
This study addresses this difficulty by providing different classifications of the literature based on a variety of research criteria.
Originality/value – This study is a first attempt to organize the outsourcing literature using statistical as well as decision support tools. Using cluster
analysis and discriminant analysis to explore the relationships among the contents of the studies is a new approach.
Keywords Outsourcing, Publications, Multivariate analysis, Decision support systems
Paper type Literature review
Outsourcing is a common practice among both private and
public organizations and is a major element in business
strategy. Perhaps most organizations now outsource some of
the functions they used to perform themselves. Due to
widespread outsourcing practices, it has become a frequent
topic in the literature. Numerous reasons why outsourcing is
initiated have been identified by researchers. Organizations
may expect to achieve many different benefits through
successful outsourcing, although there are significant risks
that may be realized if outsourcing is not successful. There is
an abundance of outsourcing literature where many benefits,
risks, motivators, and decision factors have been presented
although the relationships, commonalities and disparity
among the contents of these studies have not been
The purpose of this study is twofold. First we review the
outsourcing literature with the objective of identifying those
references that may provide guidance for managers and
researchers. The review of the literature is organized based on
the outsourcing decision framework given in Figure 1
(Kremic and Tukel, 2003). The figure depicts the typical
elements of the outsourcing decision and shows where the
motivators, benefits, risks, and factors are typically
encountered in such decisions.
Second, the studies in the literature are analyzed based on
their content. The aim is to categorize and identify
relationships among the studies which are not apparent
when the studies are considered individually. The topics
discussed in the studies are described by a set of variables and
then statistical procedures are applied. The intention here is to identify which topics are commonly discussed together as
well as the combinations which seldom appear. In addition,
we attempt to form groupings of the studies based on their
content, with the objective of identifying areas that need
The organization of the paper is as follows. We first review
the reasons that outsourcing is initiated. What are the
outsourcing motivators that are identified in literature? This
question is addressed in the second section and may help
managers determine if outsourcing is an appropriate option in
their situation. The next sections are devoted to identifying
the benefits and risks that are commonly expected with
outsourcing. A detailed discussion on factors which may
impact outsourcing decisions is also provided and followed by
the multivariate analyses and the reports on the findings. The
last section presents the general conclusions and highlights
possible areas for further research.
Motivations for outsourcing
There are three major categories of motivations for
outsourcing: cost, strategy, and politics. The first two
commonly drive outsourcing by private industry. Political
agendas often drive outsourcing by public organizations
(Kakabadse and Kakabadse, 2000a). While there may be
three categories, outsourcing activities are likely to be initiated
for more than one reason and in fact, may be driven by
elements from all three categories. For example, the
outsourcing of taxing and health services for the British
government was driven by elements from both the cost and
political categories (Willcocks and Currie, 1997). The
political climate favored privatization because of the belief
that private firms are more efficient and provide better service
than the public counterparts. Cutting the cost of providing
services also drove the British government’s outsourcing
Each of the three major categories is discussed in more
detail in the following sub-sections.
Cost driven outsourcing
Much of the literature identifies the desire to save costs as an
explanation for why outsourcing occurs (Arnold, 2000; Aubert
et al., 1996; Bienstock and Mentzer, 1999; Bergsman, 1994;
Brandes et al., 1997; Fan, 2000; Kriss, 1996; Laarhoven et al.,
2000; Vining and Globerman, 1999; Willcocks et al., 1995). In
theory, outsourcing for cost reasons can occur when suppliers’
costs are low enough that even with added overhead, profit,
and transaction costs suppliers can still deliver a service for a
lower price (Bers, 1992; Harler, 2000). One may wonder how
an organization can achieve enough savings to cover an
additional layer of overhead and still meet profit requirements
yet perform a function for less than another organization
already doing the function. Specialization and economies of
scale are mechanisms used to achieve this level of efficiency
(Klainguti, 2000; Ashe, 1996; Kakabadse and Kakabadse,
2000a; Quinn et al., 1990a, b; Roberts, V. 2001). In fact, cost
savings due to outsourcing can be quite significant. In a survey
of 7500 public organizations in Australia, the outsourcing of
cleaning services saved an average of 46 percent over in-house
performance of the service (Domberger and Fernandez,
A desire to save indirect costs may also drive outsourcing.
Having fewer employees requires less infrastructure and
support systems (Fontes, 2000; Hubbard, 1993) which may
result in a more nimble and efficient organization. Some
organizations outsource to achieve better cost control (Alexander and Young, 1996; Sheehan, 1993) while others try
to shift fixed costs into variable costs (Anderson, 1997;
Chemical Week, 2000). These are just a few examples of the
potential savings that organizations are hoping to realize with
Although organizations may outsource for cost related
reasons, there are no guarantees that expected savings will be
realized. There is increasing evidence that cost savings have
been overestimated and costs are sometimes higher after
outsourcing (Bryce and Useem, 1998; Cole-Gomolski, 1998;
Pepper, 1996); Vining and Globerman, 1999; Welch and
Nayak, 1992). As an example, again in the survey by
Domberger and Fernandez mentioned above, the outsourcing
of IT resulted in an average 9 percent increase in costs
(Domberger and Fernandez, 1999).
In addition to not realizing the costs that originally drove
the outsourcing initiative, there are also some additional
indirect and social costs that may be incurred (Gillett, 1994),
(Maltz and Ellram, 1997). Indirect costs may include contract
monitoring and oversight, contract generation and
procurement, intangibles, and transition costs. Capital
expenses incurred by the relationship should also be
calculated (Hubbard, 1993; Bounfour, 1999; Burzawa,
1994; Cole-Gomolski, 1999; Kakabadse and Kakabadse,
2000a; Vining and Globerman, 1999).
The social costs of outsourcing may be difficult to quantify
but they can be significant. Outsourcing may result in low
morale, high absenteeism, lower productivity, etc. (Eisele,
1994; Kakabadse and Kakabadse, 2000a; Walsh, 1996).
Further the social costs are not necessarily limited to the
organization. Lafferty’s and Roan’s (2000) study suggests that
the education and skill level of a whole class of workers may
be declining due to outsourcing of public services.
Contractors are less willing to pay for employee education
and development.
The message in the literature is that the desire for cost
savings may drive many outsourcing initiatives. The literature
shows that significant savings can result. However, savings are
not a given. Apparently the effects of outsourcing on an
organization’s cost are not yet fully understood and perhaps
the variables and their relationships are more complex than
Strategy-driven outsourcing
More recently the main drivers for outsourcing appear to be
shifting from cost to strategic issues such as core competence
and flexibility (DiRomualdo and Gurbaxani, 1998; Elmuti
and Kathawala, 2000; Harris and Giunipero, 1998; Lankford
and Parsa, 1999; Meckbach, 1998; Muscato, 1998; Mullin,
1996; Quinn, 1999; Roberts, V. 2001; Wright, 2001). In
general, the literature supports outsourcing as a strategy,
which may offer improved business performance on
numerous dimensions (Brandes et al., 1997; Dekkers, 2000;
Klopack, 2000; McIvor, 2000b; Moran, 1997; Old, 1998;
Prahalad and Hamel, 1990; Quinn et al., 1990a, b). Perhaps
the most often cited strategic reason for outsourcing is to
allow the organization to better focus on its core competencies
(Sislian and Satir, 2000; Quinn and Hilmer, 1994; Quinn,
1999). Because of intense competition, organizations are
forced to reassess and redirect scarce resources (Works
Management, 1999; Drtina, 1994; Jennings, 1997; Ketler
and Walstrom, 1993; Kriss, 1996; Leavy, 1996; Ngwenyama
and Bryson, 1999; Quinn, 1999; Razzaque and Chen, 1998).
Resources are typically redirected to where they make the
greatest positive impact, namely the organization’s core
In addition to refocusing resources onto core competencies,
other strategy issues which encourage the consideration of
outsourcing are restructuring, rapid organizational growth,
changing technology, and the need for greater flexibility to
manage demand swings (Eisele, 1994; Iyer and Kusnierz,
1996; Kakabadse and Kakabadse, 2000a; Lankford and
Parsa, 1999; Large, 1999; Livingston, 1992; Pinnington and
Woolcock, 1995). Flexibility appears to be an important
driver not just from a scale perspective but also regarding the
scope of product or service. Organizations need to react
quicker to customer requirements and outsourcing is seen as a
vehicle to accomplish this. Outsourcing may also be perceived
as a way to reduce the organization’s risk by sharing it with
suppliers and at the same time acquire the positive attributes
of those suppliers. The partnerships that result from
outsourcing may enable an organization to be a world-class
performer for a whole suite of products and services where it
could only be an average performer by itself. This strategy
results in a so-called “virtual organization” where functions
are outsourced to multiple vendors under one agreement.
Together the suppliers perform an integrated set of services.
There are, however, potential pitfalls when outsourcing for
strategic reasons. Organizations may “give away the crown
jewels” if they are not careful (Gillett, 1994). IBM is used as a
frequent example of a company that outsourced the “wrong”
things (the operating system). If organizations outsource the
wrong functions they may develop gaps in their learning or
knowledge base which may preclude them from future
opportunities (Earl, 1996; Prahalad and Hamel, 1990). In a
study of the aeronautics industry Paoli identifies a limit of the
virtual organization concept (Paoli and Prencipe, 1999).
Specifically, in highly integrated and evolutionary
technologies, applying the traditional core competence tests
may result in outsourcing too many or the wrong functions.
Literature also indicates that in industries with complex
technologies and systems, internal synergies may be lost when
some functions are outsourced. This could result in less
productivity or efficiency among the remaining functions
(Quinn and Hilmer, 1994).
Politically-driven outsourcing
There are several reasons why a public organization may
behave differently than a private firm and therefore may have
different outsourcing motivators. For example, Avery (2000)
argues that the performance of a service by the public
laboratory is not based on market demand or profitability.
The issues may be more social than economic. He uses the
example of the public organization detecting a virus or health
hazard, whereas the private organization would be in the
business of treating the infected for a fee. Even when the
services appear to be identical, the products may be very
different. Industry performs a service to make money whereas
the public organization attempts to ensure general well being;
a different goal and mission. So while cost and strategy may
drive private firms, the desire for the general well being of
citizens may drive outsourcing by public organizations.
Other factors that may be drive outsourcing by public
organizations include the agendas of elected officials, public
opinion, and current national or international trends (Avery,
Because public organizations are sometimes perceived as
inefficient and bureaucratic, political candidates may promote
outsourcing ideas, particularly at election time, to
demonstrate their willingness to make positive changes in
the district. Once laws are enacted, the public organization
has no choice but comply. In such situations the outsourcing
drivers are the governing laws and executive orders; another
recognized reason for outsourcing by public organizations
(Kakabadse and Kakabadse, 2000a).
Yet another reason for public sector outsourcing may be
better accountability. Deakin and Walsh (1996) find that
managers in public organizations generally realize an
accountability improvement in the particular function being
outsourced. However, the managers also believe that there is a
simultaneous decline in accountability to the public. The
explanation is that a supplier works for the government and
performs the functions to satisfy the government
representative whereas a government employee works for
the public and keeps their interests primary.
Willcocks and Currie (1997), and Willcocks et al. (1995)
write on information technology (IT) outsourcing and find
that in public organizations one of the four primary drivers for
outsourcing IT is the bandwagon effect. Apparently operating
“like a business” has appeal for the public organization. The
authors also identify manager’s preference to divest of
troublesome functions as another major reason to outsource.
In summary, there is enough evidence in the literature to
suggest that outsourcing by public organizations may be
initiated for reasons quite different from private industry.
While the reasons may be different, the desired benefits are
often similar. The expected benefits are discussed in the next
Expected benefits of outsourcing
The rapid growth of outsourcing suggests that both public
and private organizations expect benefits from outsourcing.
Naturally different organizations in different circumstances
will expect different benefits. For example, all organizations
may expect costs savings even though in government
outsourcing, the typical cost savings are only about half of
what the private sector achieves (Kakabadse and Kakabadse,
2000a). It is impossible to exhaustively list every conceivable
benefit but many of the desired benefits are general enough
that they are shared across organizations. Rather than
discussing potential benefits individually in detail, they are
summarized in Table I along with a list of references.
As the table shows, the expected benefits of outsourcing
may include realizing the same or better service at a lower
overall cost, increased flexibility and/or quality, access to the
latest technology and best talent, and the ability to re-focus
scarce resources onto core functions. For the political
organization, additional expected benefits may include
better accountability and management, and a better political
posture. There also appears to be an expected benefit of
mimicking competitors or “getting rid” of troublesome
functions (Willcocks and Currie, 1997).
Potential risks of outsourcing
The literature also discusses numerous risks associated with
outsourcing. Because outsourcing is a rather recent tool of
managers the complete costs are not yet known, which posses
a risk in itself. The literature warns that there is an initial
tendency to overstate benefits and that the suppliers are likely
to perform better in the beginning of a contract to make good
first impressions (Schwyn, 1999).
The lack of methodology is believed to cause some
outsourcing failures (Bounfour, 1999; Lonsdale, 1999).
This thinking is supported by Lonsdale who suggests that
outsourcing failures are not due to an inherent problem with
outsourcing but rather the lack of guiding methodology for
managers (Lonsdale, 1999). Another difficulty encountered
with outsourcing, particularly in the US (GAO, 1997), is the
lack of skills within public organizations to manage and
monitor outsourced functions. While not discussed in detail,
(Earl, 1996) identifies 11 risks with outsourcing IT; many of
them have applicability to the outsourcing of other functions
as well.
While it is recognized that all the potential risks of
outsourcing are not currently known, an attempt is made to
identify some of the known risks in Table II.
The outsourcing literature referenced in the table warns of
the following potential risks: unrealized savings with a
potential for increased costs, employee moral problems, over
dependence on a supplier, lost corporate knowledge and
future opportunities, and dissatisfied customers. It is also
noted that outsourcing may fail because of inadequate
requirements definition, a poor contract, lack of guidance in
planning or managing an outsourcing initiative, or because of
poor supplier relations.
Potential factors to consider
In addition to benefits and risks, outsourcing literature also
discusses factors which may impact outsourcing decisions.
The factors are discussed individually in the following
paragraphs. The factors are grouped into four categories,
strategy, cost, function characteristics, and environment.
Core competence is a strategic factor that has attracted
much attention and is often linked to the outsourcing
decision. Core competence is what an organization uses to
sustain a competitive advantage. Core competencies in turn
are utilized by core functions. There is debate in literature as
to exactly what a core function is but it is widely recognized
that how core a function is should have bearing on whether or
not to outsource it (Quinn, 1999; Drtina, 1994; Jenster and
Pedersen, 2000; Quinn, 2000; Large, 1999; Lankford and
Parsa, 1999; Kakabadse and Kakabadse, 2000a; Prahalad and
Hamel, 1990; Dekkers, 2000; Elliott and Torkko, 1996;
Brandes et al., 1997; McIvor, 2000a). Quinn suggests that
“those activities – usually intellectually-based service
activities or systems – that the company performs better
than any other enterprise” are core (Quinn, 1999). In general,
a function that is more core to the organization is less likely to
be outsourced.
A second strategy factor is critical knowledge. There are
some functions in an organization that may not in-and-of
themselves be “core” but the unique data or technology they
generate and feed into other processes is critical. Similarly,
there are functions in organizations which generate data or
knowledge that the organization wants to be cognizant of and
in control of. The critical knowledge factor is intended to
describe this type of function. In general, if a function
provides critical knowledge it is less likely to be outsourced.
Lack of internal human resources is another factor
identified within the strategy category. Public organizations
may be particularly impacted by lack of resources. They are
historically more restricted in their hiring and termination
practices than private-sector organizations. There are often
strict guidelines on the number of civil servants that can be
employed. When public organizations are restricted from
hiring employees to replace those retiring or exiting, there is
more workload for those remaining. Further, employees that
left took their knowledge and skills with them, leaving a
void in the organization. The organization must make
strategic decisions about how to re-locate the workforce that
remains. There may be cases when the best alternative for
the public organization is to acquire the needed skills from
outside sources. In both public and private firms access to
the people with specialized skills may be an issue. In
general, a function is more likely to be outsourced if there is
a lack of internal human resources to perform it (Green,
The impact on quality is the next strategy factor to
consider. The quality of an organization’s services establishes
reputation and can create demand. If the organization is
currently recognized in the industry for a high level of quality
then there may be concern by decision makers or customers
that outsourcing the function could harm quality. However, if
the organization’s quality is not held in high regard, then
outsourcing the function may be seen as a potential
improvement. Therefore quality is a relevant factor and can
be either a positive or a negative influence on outsourcing
(Anderson, 1997).
Flexibility is the last factor identified in this category.
Flexibility, as discussed here, is intended to include demand
flexibility, operational flexibility, resource flexibility, or the
flexibility of a number of other strategic elements. Like
quality, flexibility can be impacted positively or negatively by
outsourcing. Long contracts outsourced into a limited market
have sometimes resulted in a loss of flexibility (Antonucci
et al., 1998; Bryce and Useem, 1998). However, large bureaucratic organizations may improve on their flexibility by
outsourcing. According to the literature, organizations
sometimes consider outsourcing in an effort to increase
The next factor category is cost. There is some literature
that suggests that most outsourcing is primarily motivated by
the organization’s efforts to reduce costs (Meckbach, 1998;
Hendry, 1995; Welch and Nayak, 1992). If a function is to be
outsourced for cost reasons, then it is assumed that the
current in-house costs are higher than the expected costs for
purchasing the service. However, there is significant
uncertainty about the expected savings generated by
outsourcing. Cost savings may not be as high as sometimes
reported. Literature also suggests that determining accurate
in-house costs to compare to can be difficult. Despite the
uncertainty, many organizations outsource to reduce costs
and therefore the higher the internal cost to perform the
function relative to the expected cost of purchasing the service
the more likely the function is to be outsourced.
The next factor category relates to the characteristics of the
functions themselves. Some functions simply lend themselves
better to being outsourced. Function characteristics, such as
complexity, degree of integration, structure etc., unlike the
strategy or cost factors, are generally not unique to an
Several factors have been identified that fall into this
category, the first being complexity. Complexity refers to the
difficulty of recognizing or understanding the variables and
the interactions that surround a function. An example of a
complex function may be basic research in any of the sciences.
Often it is not known what is being sought and therefore there
may not be a specific expected outcome. Because outsourcing
is the purchasing of a service under contract it generally
assumes that both the outsourcer and supplier know what is
to be delivered and under what terms (Prencipe, 1997). It
may be more difficult to articulate the requirements and terms
for complex functions. It also requires a greater investment by
a supplier to learn to perform complex functions. In general,
the more complex a function is the less of a candidate it is for
Integration is another function characteristic that influences
the outsourcing decision. Integration refers to the degree the function is linked into other functions and systems within the
organization. The more integrated the function is the more
interactions and communication channels there are to
maintain and monitor. It is often difficult to achieve
effective communication and coordination within an
organization. Moving a function across organizational
boundaries and perhaps adding several new layers of
departmental boundaries won’t make the task any easier.
Differing organizational cultures and goals may also impede
successful interaction with the relocated function. Therefore a
function that is heavily integrated is less of a candidate to
outsource (Prencipe, 1997; Paoli and Prencipe, 1999).
The next function characteristic is asset specificity. This
describes the case where durable equipment or products are
generated by the outsourcing arrangement and they have little
value outside of that function. Asset specificity is an
outsourcing issue because the supplier has little incentive to
put resources into maintaining or upgrading the durable items
because they have no value for him apart from the agreement.
Ultimately the supplier has greater leverage to charge higher
rates. In general, the greater the asset specificity the less of a
candidate the function is for outsourcing.
The structure of a function also impacts the decision to
outsource it. Structure relates to the degree the function
follows a predictable pattern; a pattern which may be defined
in a checklist. The more structured a function is the easier it is
for a less experienced person to perform it with proper
instructions. A more structured function is a better candidate
for outsourcing.
The number of employees impacted is likely to influence
the decision to outsource a function. Displaced employees are
a sensitive issue, whether the organization’s goals are to
displace as many as possible or as few as possible. If an
organization is seeking to divest of employees, then a function
utilizing a relatively large number of personnel will be
attractive for outsourcing; however that same function may be
an unlikely candidate if the organization wants to minimize
the number of employees impacted. The determination
whether the number of employees impacted has a positive or
negative effect on the outsourcing decision is made on a case
by case basis.
The final category of factors relate to the internal and
external environment faced by the organization.
For organizations with unique missions or specialized skills
there may be only a few if any outside suppliers possessing
those skills. An example is maintenance of the Hubble Space
Telescope. There are not many organizations that could
assume this function. A function that requires skills that are
difficult to find externally is less likely to be outsourced
(Antonucci et al., 1998).
Political pressures have undeniable influence on public
sector organizations. As mentioned in the second section,
public organizations don’t necessarily make decisions based
on cost and profit. The agendas of elected officials, public
opinion, and current national or international trends may all
influence the actions of the public organization. There may be
certain functions or types of functions that are particularly
visible in the public eye and therefore may receive increased
outsourcing pressure. Airport security is a recent example of a
function being pulled back in-house by the government.
The internal political environment may also influence the
outsourcing decision. The opinions of influential people
within the organization may have bearing on decisions even
though they may not have any formal outsourcing decision
authority. The perceptions of employees, unions, or union
leaders may also influence whether a function should be
Another environmental factor is the preference of the
managers that do have formal influence on the outsourcing
decision. For example, a manager may favor a certain
function and want it to stay in-house. That function is less
likely to be outsourced than one that doesn’t have the
manager’s preference. The degree to which preferences
influence a decision may be difficult to predict. Further, the
influence may be hidden in supporting documents or ancillary
decisions and thus may be difficult to identify. None-the-less
manager preference is an environmental factor to consider.
The fifth environmental factor identified is the legal
environment. Legal factors relate to the degree the
candidate functions are “tied up” in current legal
arrangements. Legal factors may be union agreements,
contracts with other suppliers, or other rules and regulations
that govern the performance of a function, such as veteran
preferences and minority initiatives. In general, the more legal
hurdles to overcome the less likely the function is to be
The next environmental factor identified is the actions of
competitors. Willcocks and Currie document that one of the
reasons that many firms try outsourcing is because others are
doing it (Willcocks and Currie, 1997). The unwritten
assumption appears to be that if the competitors are doing
it, it must be good. In general, if the organization’s
competitors are actively outsourcing a function it is more
likely to be outsourced.
The potential for conflict of interest is another
environmental factor to consider. The outsourcing of some
functions may result in a situation where the supplier has to
act contrary to their other interests. An example is when a
supplier is placed into a role where they could create work for
themselves. Consider a supplier that has been given the
responsibility of strategic planning and recommending of IT
systems for an organization. If that supplier also made or
developed IT systems, then there may be speculation by the
organization when the supplier recommends its own products
for upgrades. The question may be asked: “Are the upgrades
really needed or is the recommendation based more on a
desire to generate additional business”? In general, functions
are less likely to be outsourced as the potential for conflicts of
interest increases (Graham, 1996).
The degree of uncertainty is the last factor. If the
environment or disposition of a function is highly uncertain
it may be more difficult to successfully outsource that
function; especially at a fair rate into a firm long-term
contract. When the inputs, requirements, or costs associated
with a function are uncertain, a potential supplier will have a
difficult time assessing what is a fair price to charge.
Consequently, they will demand a higher rate to assume the
extra risk. In addition to higher rates, there are likely to be
more problems with such functions during the course of the
arrangement. Greater uncertainty may also make it more
difficult to define the requirements and expectations. In
cultures where formal arrangements are the norm, loose
definition often results in change orders, unexpected costs,
and sometimes a negative impact on relationships. In general,
successful outsourcing of highly uncertain functions is more
Multivariate analyses of the literature
In this section we report on the cross-tabulation and the
cluster analyses which were conducted on the outsourcing
literature. Both methods are used for categorizing the
literature and highlighting disparity and trends in the results
presented in these studies.
The database
A total of 210 studies in the last two decades are included in
the database, with well over half being between the years 1996
to 2000. The studies in the database appeared in academic
journals and e-journals. The keywords in the outsourcing
decision framework provided in the second section are used in
the search process.
Based on the forgoing discussion and Tables I and II, a total
of 29 variables are identified and each paper is classified based
on whether a particular variable is present or not. The
variables describe the studies and are derived from the
decision framework presented in Figure 1. Seventeen of the
29 variables relate to outsourcing benefits, risks, motivations
and factors. Five variables relate to outsourcing benefits: does
the study discuss cost savings, or focus on core competence,
flexibility, access to technology, and access to skills? Another
five variables relate to risks associated with outsourcing: loss
of core competence, supplier problems, employee morale, loss
of skills, and unrealized savings. Motivation to outsource is
represented by three variables: cost driven, strategy driven,
and politically driven. Factors which may impact the decision
of which function to outsource are represented by four
variables: strategy, cost, function characteristics, and business
environment. The rest of the variables provide additional
descriptions of the studies. Three of them are used to define
the location of the study: does the study discuss outsourcing
practices in Europe, in the USA, or in other regions? Another
three variables are for the type of study: case, theoretical, and
decision support tool. The type of organization is also defined
by three variables: for-profit, non-profit, and governmental
organizations. Finally four more variables are included in the
analysis to further describe the studies: human resources
issues related to outsourcing, the impact of knowledge
management, general core competence issues, and the
impact of R&D strategy on outsourcing. For each study
each of these variables can have a value of zero or one
depending on whether that study deals with the
corresponding variable. This results in a matrix with a total
of 210 £ 29 entries.
Cross-tabulation analysis
In order to combine and compare the information provided in
the database and thus identify the relationships between the
contents of the studies we first constructed pivot tables using
Microsoft EXCEL software package. A pivot table report is a
means of organizing the data with the purpose of identifying
trends and commonalities. Using the 29 variables as pivot
table fields the cross tabulations are constructed and the
counts of studies that satisfy both conditions on various
corresponding variables are determined. The significant
findings of the tables are as follows:
. Of the 210 studies 80 of them (38 percent) specify the
type of organization, with 56 percent addressing for-profit
. Of the 85 studies that state cost savings as the main
outsourcing benefit, 57 percent of them refer to for-profit
organizations compared to only 2 percent for non-profit
. No studies in the database indicate that non-profit
organizations outsource in order to access technology or
to access skills and talent.
. There are twice as many studies which discuss supplier
problems as those which address unrealized savings. Both
risk categories are more commonly associated with
governmental organizations as opposed to for-profit
. Of the 85 studies which refer to cost reduction as a main
driver for outsourcing, 53 percent of them are aimed at for
profit organizations whereas for the 95 studies with
strategy as the main driver 58 percent of them are aimed
at for-profits. On the other hand, 75 percent of the 12
studies which discuss political motivations refer to
governmental organizations.
. Only one third of the 32 studies which discuss cost being a
major factor that affects the decision of which functions to
outsource refers to for-profit organizations while two
thirds refer to governmental organizations. Note also that
out of a total 33 governmental studies only 12 of them
study the outsourcing efforts of US governmental
. Of the total studies in the database, 79 percent are of
theoretical type, and 56 percent of them refer to for-profit
organizations, while only 2 percent of them investigate
non-profits. There are a total of 18 case studies in the
dataset with 62 percent considering for-profits and 38
percent dealing with governmental organizations.
. All three types of studies (case, theoretical, and decision
support) concentrate on US organizations.
. The studies discussing outsourcing benefits are twice as
likely to refer to the US practices compared to European.
The largest difference occurs when the expected benefit is
access to skills. The number of studies referring to US
practices is four times those of Europe. Among the
numerous risks associated with outsourcing, the supplier
problems are discussed more often when US practices are
considered. The risk of unrealized savings is discussed
more often in the European studies.
. When the motivation for outsourcing is cost driven (85
studies), 80 percent of them report cost savings as a
benefit that they seek. In addition, 81 percent of the 37
studies which discuss the risk of not realizing the expected
savings also discuss cost savings as an outsourcing benefit.
Chi-square test
The second analysis is a chi-square testing of the relationship
between the variables in the database. For the purposes of the
chi-square test we defined several binary variables: B, R, M,
and F. B refers to whether or not any of the five benefit
variables are discussed in the study. Similarly R, M, and F
refer to whether or not any of the risks, motivations or factor
variables are discussed.
The results are presented in Table III. The chi-square tests
on the pair-wise associations are all highly significant and
positive, indicating that papers which discuss one of the B, R,
M, F tend to also discuss each of the others. For example,
when a paper discusses outsourcing benefits it is more likely
that it also discusses risks and on the other hand if it does not discuss benefits, then it is also less likely that it will discuss
risks. In the table only the statistically significant relationships
are presented. The first row, which corresponds to benefits
(B) and the second column corresponding to risks (R)
contains the count of the studies for each possible
combination of the two variables. For example, 56 studies
discuss both benefits and risks while 47 studies discuss
benefits but not risks. The value of the chi-square statistic for
testing the independence of B and R is then 18.1 which is
highly significant indicating that when a study discusses
benefits it also tends to discuss risks. This pattern applies to
all the other pairs in the table.
Next, we performed chi-square tests comparing the rest of
the variables to B, R, M, F. Most of the relationships were not
statistically significant. Among the significant ones are
Knowledge Management (KM) with each of B, R, M, F,
and Decision Support (DSS) with each of B, R, and M. In
each case here the relationship is strongly negative. That is,
when knowledge management is discussed then it is less likely
that any of B, R, M, F is discussed. Similarly, when the study
type is decision support it is much less likely that any of B, R,
M is discussed. We find several other significant relationships
involving the variables Government, For-profits and Location
(USA, Europe). Government has a positive relationship with
B and R, for-profits has a positive relationship with B, US has
a positive relationship with B while Europe has a positive
relationship with R.
Note that the results of chi-square testing provide statistical
support to the findings based on the pivot table analysis.
Cluster analysis
In the foregoing analysis we examined the relationships
between the variables describing the content of the papers.
Next, we will group the papers based on these variables.
For this purpose we use cluster analysis available in SAS
(SAS, 1999). Each study is considered to be a point in
the multi-dimensional space determined by the values of the
variables. These points are then clustered, based on the
distance between them. Using the MODECLUS procedure in
SAS we were able to choose parameters which lead to
different numbers of final clusters. Although the possible
number of clusters varied from three to several dozen we
chose the six cluster solution as providing a suitable balance
between an overly detailed classification and an overly
aggregated one.
Portrayal of clusters
Stepwise discriminant analysis was then used to investigate
which variables are important in determining the membership
of the clusters. The stepwise discriminant analysis is
commonly used to evaluate differences among groups of
observations. We used the SAS STEPDISC procedure which
also provides the relative importance of each of the variables
in predicting group membership. Table IV summarizes the
important variables for each of the clusters and how the
clusters relate to them, i.e. a positive, or a negative
correlation. The variables in the tables are ranked from the
most important (in terms of the value of the F statistic) to the
least important in forming the cluster. The largest cluster
contains 127 studies and is determined by values on ten
variables. Note that the small clusters are characterized by
strong relationships with a few variables while the large cluster
is characterized by variables that are not addressed in the
The common characteristics of cluster 1 are that the studies
discuss decision support systems and are theoretical. However
cost as a motivator for outsourcing is not discussed nor is
knowledge management or functional characteristics as a
factor in deciding what to outsource. In cluster 2 studies
discuss both knowledge management and human resources
issues without linking it to a specific location. Furthermore
they are not case type studies. The primary common
characteristic of studies in cluster 3 is that they discuss
access to technology as a benefit of outsourcing and loss of
employee morale as a risk. The cluster also tends to contain
studies addressing non-profit organizations as well as those
discussing access to skilled workers as an outsourcing benefit.
Cluster 4 consists of case studies where strategy as a factor
affecting what to outsource is not discussed. In Cluster 5
function characteristics as a factor is the common topic while
risk of unrealized savings and risk of loss of knowledge are
The relationship with the primary variables determining
membership in cluster 6 is negative. This indicates that
studies in this cluster tend not to discuss decision support
tools or access to technology as a benefit or any of the human
resources issues. However, they do tend to discuss cost driven
motives, core competence issues related to outsourcing and
the outsourcing practices taking place in Europe and other
locations but not the USA.
Multivariate analysis of the literature is instrumental in
identifying commonalities, patterns, and gaps in the
outsourcing literature. In general outsourcing studies are of
theoretical type, discussing benefits, risks and motivators for
for-profit, US organizations. European studies focus more on
risks related to outsourcing and address governmental
entities. The studies addressing knowledge management
tend not to cover benefits, risks or motivators related to outsourcing, a pattern which holds for studies addressing
decision support issues also.
It is also important to note that most studies do not
consider the distinct position of non-profit organizations
when it comes to outsourcing practices. Non-profit
organizations do not typically have cost reduction as a
primary objective. They are typically politically driven. This
observation should be a motivation for researchers to expand
the treatment of non-profits. The cluster analysis indicates
that approximately 40 percent of the studies can be
characterized based on a strong relationship to relatively few
of the variables. These variables usually do not relate to
motivators, type of organization, or location.
Organizations are doing more outsourcing than ever before
and managers are in desperate need of information in an
organized form that will help them identify opportunities,
challenges, and decision factors related to outsourcing. There
is an abundance of information related to outsourcing in the
literature that is waiting to be put into a more structured form
for better decision support. With this study we attempt to
accomplish this task. A comprehensive list of recent
outsourcing studies is presented and analyzed statistically
based on their content. The analysis helped identify emerging
paths as well as less visited areas in the literature. Our general
observation based on this analysis is that literature is rich in
terms of presenting the possible benefits, risks, and strategic
issues to outsourcing. However, when it comes to offering
tools and guidelines in terms of decision support, the
literature is lacking and needs additional work.

No comments: