Post on 13-May-2023
1
Why Distance Matters: The Dynamics of Offshore Location Choices
Stephan Manning Duke University, The Fuqua School of Business
1 Towerview Drive, Durham, NC 27708, USA
Email: sdm24@duke.edu
Marja Roza Erasmus University, Dep. of Strategic Management and Business Environment
P.O. Box 1738, 3000 DR Rotterdam, The Netherlands
Email: MRoza@rsm.nl
Arie Y. Lewin Duke University, The Fuqua School of Business
1 Towerview Drive, Durham, NC 27708, USA
Email: AYL3@duke.edu
Henk W. Volberda Erasmus University, Dep. of Strategic Management and Business Environment
P.O. Box 1738, 3000 DR Rotterdam, The Netherlands
Email: h.volberda@rsm.nl
Duke CIBER Working Paper
2
Why Distance Matters: The Dynamics of Offshore Location Choices
Abstract
Offshoring of technical and administrative work has become an established business practice in
recent years. One key aspect of offshoring decisions is the choice of location. Prior research
suggests that availability of expertise and low cost labor are major location choice factors,
whereas geographical distance does not matter much because of advancement of IT. By contrast,
we show that access to talent pools, external expertise and geographical proximity are important
trade-offs and part of a competitive dynamic that unfolds over time. Based on comprehensive
data of early and more recent offshore investments of U.S. and Dutch companies, we conduct a
multi-level analysis of the likelihood of nearshore vs. farshore implementations over time.
Findings suggest a co-evolutionary dynamic of location choices, changing industry and labor
market conditions. Results point to limitations of conventional theories of internationalization
and may inspire follow-up studies accounting for these multi-level dynamics.
Keywords: Offshoring, Nearshoring, Location Choice, Co-evolution, Search for Talent
Introduction
The relocation and sourcing of business functions and processes outside national borders – also
called: offshoring – has become an established business practice (Doh, 2005; Kenney et al.,
2009; Manning et al., 2008). Since the beginning of the 1990s, in particular U.S. companies have
engaged in offshoring administrative and technical work, including IT, finance & accounting,
and legal services, but also more advanced functions, including software and product
development. In more recent years, an increasing number of European companies have also
started to relocate various business functions abroad, in particular to emerging economies.
Reducing labor costs and accessing growing pools of qualified personnel outside their home
countries have been major offshoring drivers (e.g. Lewin & Couto, 2007; Lewin et al., 2009).
3
One key aspect of making offshoring decisions is the choice of the offshoring location.
The rationale for selecting particular locations has attracted a number of studies in recent years
(e.g. Doh et al., 2008; Bunyaratavey et al., 2007; A.T. Kearney, 2004). Most of these studies
stress the importance of labor cost advantages, level of education and expertise, as well as
language capabilities as location selection factors, but they also list a number of risks associated
with locations, such as political stability, wage inflation, and protection of intellectual property
(see e.g. Doh et al., 2008; A.T. Kearney, 2004; Lewin & Couto, 2007). Moreover, some recent
studies suggest that over time specialized geographical IT and knowledge service clusters have
emerged, e.g. in Russia, Eastern Europe, India and China, that attract function-specific
investments (e.g. Bresnahan et al., 2001; Dossani & Kenney, 2007; Manning et al., 2008).
Interestingly, however, most studies on offshore location strategies neglect the
importance of geographical distance. Many scholars believe that distance does not matter much
in service offshoring (e.g. Blinder, 2006). The main argument is that advanced ICT has
facilitated the modularization and reorganization of tasks and has made long-distance
coordination and communication less costly (e.g. Kenney et al., 2009; Metters & Verma, 2008).
However, empirical evidence, e.g. data collected by the Offshoring Research Network (ORN),
suggests that distance does matter in offshore investment decisions: A significant number of
companies, in particular from Western Europe, have preferred nearshore locations for their
offshore investments. Nearshoring means that captive or third-party offshore implementations
are made relatively close to the home country. By contrast, U.S. companies have mostly
offshored to farshore locations, such as India, China or the Philippines. Only recently, more
European companies seem to have shifted operations to more distant locations.
4
We seek to investigate preferences for nearshore vs. farshore locations in greater detail as
well as changes over time, based on comprehensive data of U.S. and Dutch companies, collected
by ORN. Thereby, we take a dynamic, multi-level analytical approach by looking at influences at
the macro-economic, population, firm and task level. We propose, based on our data, that a co-
evolutionary perspective (see Lewin & Volberda, 1999; Hutzschenreuter et al., 2008) helps
understand the dynamics of offshore location preferences over time (see also Manning et al.,
2008; Lewin et al., 2008). Our findings may not only stimulate future research on offshoring
strategies and location choices. They also shed light on the dynamics of internationalization in
more general. We argue that existing – rather market-oriented – theories of internationalization,
e.g. the OLI framework or the Uppsala model, are insufficient in explaining these recent
offshoring dynamics. A co-evolutionary model can be a promising alternative starting point for
analyzing more recent and future trends of internationalization.
Offshore Location Choice and the Role of Geographical Distance
The choice of location is a key dimension in making offshoring decisions (see e.g. Doh et al.,
2008; Lewin & Couto, 2007; Bunyaratavey et al., 2007). In more general, offshoring refers to the
(re-) location of business processes and functions outside the borders of the home country (e.g.
Manning et al, 2008; Kenney et al., 2009). Offshoring is often confused with outsourcing, which
denotes the transfer of tasks and processes to external providers. Offshoring, however, may
include both captive models (setting up wholly owned subsidiaries) and offshore outsourcing
(engaging external international and local providers to deliver particular services). Also,
offshoring must not be confused with market-oriented foreign investment, e.g. setting up sales
operations. Rather, offshoring is a sourcing strategy designed to set up local operations
5
supporting domestic and global activities of the firm (see Manning et al., 2008). With regard to
offshoring of technical and administrative tasks, e.g. IT, finance and accounting, software and
product development, the most important factor being ‘sourced’ offshore is qualified personnel
(see Lewin et al., 2008; Manning et al., 2008), and – if available – external capabilities and
subject-matter expertise (offshore outsourcing).
The most important offshoring drivers for U.S. and Western European companies are
saving labor costs and accessing qualified personnel. Offshore location decisions to a great
extent reflect these general drivers (see e.g. Lewin & Couto, 2007). Accordingly, labor cost
advantages, size of the talent pool available and access to subject-matter expertise are important
criteria in choosing offshore locations for particular services (see e.g. Doh et al., 2008). Along
these lines, some scholars argue that the success of India and China as preferred offshore
destinations can be mainly explained by the quantity and quality of technically skilled labor
available (see e.g. Freeman, 2006; Dossani & Kenney, 2007). However, recent studies indicate
that in particular European companies often select nearshore locations close to their home
countries rather than India or China for making offshoring implementations (see e.g. Lewin &
Couto, 2007; A.T. Kearney, 2004). Notably, a number of U.S. companies also consider Canada
and Mexico in particular as alternatives to further remote locations. We seek to understand why
this is and, yet, why recently a shift towards more remote locations can be observed.
In more general terms, the question is why and to what extent distance matters in offshore
location decisions. The role of distance has been widely neglected in the more recent offshoring
literature. Some scholars even argue that because of advances in IT, decreasing long-distance
communication costs and the emergence of specialized geographical clusters providing talent
6
and services for MNCs globally, distance should not matter much in service offshoring (e.g.
Blinder, 2006).
The more general literature on internationalization however suggests that distance can be
an important factor in any foreign investment decision. Rugman and Verbeke (2004) for example
show, based on sales data, that the majority of foreign investments is made within the region of
the home country. One important reason why many companies set up manufacturing and sales
operations close to their home countries are cost and time advantages. For example, geographical
proximity to headquarters or other central locations may reduce transportation and travel time
and costs. In the offshoring context, in particular for those tasks that require intensive training of
staff and coordination with head quarters or other locations travel costs can be expected to be an
important factor in choosing locations. Related to this, time zone differences seem to affect
coordination and communication between locations (O’Leary & Cummings, 2007). On the other
hand, strategies targeted at establishing 24h knowledge factories may favor implementations in
more remote time zones (e.g. Gupta et al., 2007).
Another equally important and historically related reason why many companies seem to
prefer geographically close destinations for foreign investment is greater familiarity with
economic, cultural and institutional conditions in countries within the region (e.g. Rugman &
Verbeke, 2004). Most prominently, Johanson and Vahlne (1977) argue that investment in
contexts characterized by ‘psychic proximity’ facilitate learning that is necessary to expand the
global footprint. Psychic proximity relates to the accessibility of information about local market
conditions. Availability of location information in the context of offshoring can reduce search
costs, e.g. costs involved in finding talent or qualified service providers, which may facilitate and
speed up the implementation of offshore projects. Reasons why geographical proximity often
7
correlates with psychic proximity (or greater familiarity of context conditions) are e.g. the
embeddedness of companies in local and regional networks, trade relationships between
countries and prior foreign investments.
Similarly, IB scholars have argued, based on Hofstede’s concept of cultural distance
(Hofstede, 1980), that companies are more likely to invest in host countries that share norms and
values with home countries (e.g. Kogut & Singh, 1988; for a critical view, Shenkar et al., 2008).
Again, geographical proximity is often correlated with cultural proximity. One important factor
here is the use of the same or a similar language which facilitates intra- and interfirm
communication and coordination. In fact, similar norms and language have been listed as
primary reasons in previous offshoring studies for the preference of many Western European
companies to select locations in other European countries (e.g. A.T. Kearney, 2004). Related to
this, Xu & Shenkar (2002) argue that similar regulative, normative and cognitive institutions (see
in general, Scott, 2001) – in short: institutional proximity – may affect location choice. Along
these lines one could argue that the fact that many European countries belong to the European
Union and therefore adhere to the same or a similar legal system reduces institutional distance in
favor of nearshore investments.
Most of these theoretical approaches, however, have two major weaknesses. On the one
hand, they may not explain differences in location preferences. For example, according to ORN
data, U.S. companies seem to have a much higher preference for India and other remote locations
than most European companies (see Figure 1). Some European companies, in turn, e.g. Spanish
companies, also seem to prefer further remote locations, e.g. in Latin America. What’s more,
companies from the same home country often have different preferences for nearshore vs.
farshore locations (see in more detail below). On the other hand, established approaches seem to
8
insufficiently explain changes in location preferences over time. As we demonstrate in more
detail below, ORN data suggests that European firms have a greater preference for farshore
locations in recent years. Figure 2, for example, shows, based on ORN data, how the distribution
of offshore investments in India by country of origin has shifted over time from U.S. to European
countries. The Uppsala school of internationalization (Johanson & Vahlne, 1977) would argue
that European firms have learned over time to establish subsidiaries in locations formerly
characterized by great psychic distance. This however neither explains why most U.S. companies
offshored to India from early on, nor does it explain why in recent years even European
companies with no or little offshoring experience are more likely to go farshore than nearshore.
----------------------------------
Insert Figure 1,2 Here
----------------------------------
We argue that a multi-level analytical approach is needed to come to grips with the dynamics of
offshore location choice. As proposed by Hitt et al. (2007), Volberda and Lewin (2003) and
others, multi-level approaches serve to simultaneously take into account influences on strategic
decisions or managerial behavior coming from multiple systemic contexts – e.g. the firm itself,
as well as macro-economic, industry, and project implementation-level factors. Also, they invite
thinking about interactions between these levels, including intended and unintended
consequences of firm-level decisions on the very contexts in which firms operate. This basic idea
has been further developed in particular by co-evolution theory which seeks to explain firm-level
adaptation-selection within changing competitive environments (Lewin & Volberda, 1999). In
the following, we outline our analytical framework in more detail. We develop hypotheses to
explain the preference for nearshore vs. farshore implementations based on explanatory factors at
9
different levels. These factors will be derived mainly from existing offshoring studies as well as
the more general IB and internationalization literature.
Explaining Nearshore Investments: A Dynamic Multi-Level Framework
To come to grips with preferences for nearshoring we apply a multi-level analytical framework.
Multi-level analyses are suitable where a phenomenon, such as a strategic firm-level location
decision, cannot be easily explained just by factors at one level, e.g. firm-level preferences,
national or industry context factors, but only by the combination and interaction between these
levels (Hitt et al., 2007). In line with recent papers on the offshoring phenomenon (e.g. Manning
et al., 2008, Lewin et al., 2009; Kenney et al., 2009) we see firm-level offshoring decisions in
general – and location decisions in particular – co-evolving with industry and macro-level trends
as well as home and host country economic conditions. The propositions we develop in the
following, however, are not meant to be exclusive, but they are designed to help understand part
of the dynamics of factors influencing location decisions in general and nearshore vs. farshore
location choices in particular. Further studies are needed to analyze those as well as other
interacting factors in greater detail.
Macro-level effects: Offshoring decisions are influenced by factors that cannot be simply related
to firm, industry or country-level trends, but that relate to trends affecting strategic decisions
across firms, industries and countries. In the context of offshoring, the advancement of ICT can
be named as one of the key factors facilitating and accelerating recent offshoring trends (Kenney
et al., 2009; Metters & Verma, 2008). In terms of location choice, the argument has been made
that advanced ICT reduces communication and coordination costs and therefore makes
10
geographical distance less relevant (Blinder, 2006). In other words, advancement of IT should
make, ceteris paribus, nearshore location decisions less likely over time. In a similar vain,
offshoring has become an established business practice, promoted by firm- and industry level
experience, but also by a broader discourse, stimulated by business press, consulting and
academia (e.g. Levy, 2005). India in particular has attracted much attention in the offshoring
discourse in recent years (e.g. Dossani & Kenney, 2007). But also China has become a much-
celebrated location for offshore investments (see e.g. Huang & Khanna, 2003). The
establishment of offshoring as an accepted business practice across industries and in particular
the fast-growing popularity of India and China should also make, ceteris paribus, nearshore
location decisions less likely over time. We therefore propose:
H1 (Macro-level effect): The more recent an implementation is (independent of company
experience), the less likely is it a nearshore implementation.
Country-level effects: One of the most important drivers for companies to make offshore
investments is access to lower-cost personnel. At the same time, we have argued, based on
previous studies, that companies prefer to make investments closer to their home countries for a
combination of reasons, e.g. search costs, coordination costs, cultural and institutional proximity.
Initially, therefore, we predict companies to make use of nearshore resources, such as qualified
low-cost personnel if available. However, in the longer term, the very offshoring motivation to
access personnel and to lower labor costs (see e.g. Lewin & Couto, 2007) will favor farshore
investments. Therefore, we argue that, ceteris paribus, the more labor resources are available
farshore vs. nearshore the more likely will a company make a farshore investment. In addition,
we argue that the lower labor costs are farshore vs. nearshore, compared to the home country, the
11
less likely will a company make a nearshore rather than a farshore investment. Both arguments
have been supported by numerous economists, e.g. Freeman (2006). Importantly, these economic
conditions may change over time. Research has shown that growing competition for talent and
wage inflation may drive companies to shift operations to other ‘second-tier’ locations. While we
cannot directly predict effects at this time, we will consider the growing competition for talent at
particular locations, e.g. hotspots, as indicators for the changing availability of lower-cost
qualified personnel over time.
H2a (Country-level effect): The bigger the pool of qualified, lower-cost personnel
nearshore (ceteris paribus), the more likely is a nearshore implementation.
H2b (Country-level effect): The bigger the pool of qualified personnel available in
farshore vs. nearshore locations, the less likely is a nearshore implementation.
H2c (Country-level effect): The lower the cost of labor in farshore vs. nearshore locations
compared to the home country, the less likely is a nearshore implementation.
Population-level effects: Offshoring location decisions are also affected by behavior of other
firms, e.g. of the same industry or national origin. With regard to offshoring, clearly companies
of the same national origin show similar patterns of decision-making (see above). Part of this
similarity stems from a common domestic economic and institutional environment. Part of it,
however, relates to imitating behavior. Case studies indicate that pioneer companies entering a
location from a particular country attract followers of the same national origin. On the one hand,
this has to do with the development of trust in local business conditions. On the other hand,
companies often actively shape the local investment environment, e.g. by building institutional
links with local administrations and universities (e.g. Manning et al., 2008). As a consequence,
12
they ‘customize’ local service capabilities (Athreye et al., 2005), attracting future investment
from companies from the same business system. From the point of view of IB, this can be
explained by the reduction of psychic and institutional distance – e.g. the increasing familiarity
of locations among firms over time (Xu & Shenkar, 2002; Johanson & Vahlne, 1977). As,
according to theory, companies tend to be initially more familiar with nearshore locations, this
dynamic is likely to positively affect the likelihood of farshore implementations.
In a similar vain, we propose that experience with locations related to offshoring
particular functions, e.g. software, product development etc., may increase the likelihood of
companies offshoring this function to a particular location. This can be also called the
specialization effect. For example, Bangalore, India, has clearly attracted mainly IT and software
development-related investment, because, over time, specialized local service capabilities have
developed that attract companies across industries (Athreye et al., 2005; Dossani & Kenney,
2007). Again, we assume that this population or industry effect is stronger for farshore than it is
for nearshore investments. In other words, when it comes to nearshore implementations,
companies can be expected to be more knowledgeable about providers and qualified personnel
based on personal experience and ‘pioneers’ are not needed so much to make nearshore
investments attractive. In sum, we predict that population experience with farshore investments
as well as specialization effects may decrease the likelihood of a nearshore implemention.
H3 (Population-level effect): The more implementations have been made farshore
relative to nearshore across firms up to a certain point in time, the less likely is a
nearshore implementation at this time.
13
Firm-level effects: Strategy research suggests that in a competitive space firms differ in terms of
strategic choices (e.g. Child, 1972), because of different resource endowments, decision-making
processes, absorptive capacity, strategic goals and managerial discretion etc. Also they respond
to economic and institutional challenges, e.g. shortage of qualified personnel or EU enlargement,
in different ways (Oliver, 1991). We therefore propose that the very factors companies perceive
to be important when making location decisions for particular offshore projects matter.
With regard to nearshore implementations, clearly the importance of geographical
proximity in a location decision can be predicted to play a key role. Importantly, the very fact
that companies invest nearshore must be separated from the importance geographical proximity
has in a particular decision. This is because nearshore investments can be potentially made for a
variety of reasons, e.g. previous investments (e.g. manufacturing), availability of special
expertise etc. Proximity itself, by contrast, is assumed to highly correlate with low coordination
costs and familiarity of the location. In this context, perceived language capabilities and cultural
proximity can be expected to be associated factors positively contributing to nearshore rather
than farshore investments (e.g. Doh et al., 2009; A.T. Kearney, 2004). These considerations,
according to the literature, need to be traded off against the perceived importance of saving labor
costs and availability of personnel in a potential offshore location. As labor costs in India and
China (and other, second tier, Asian and Latin American locations) are significantly lower than
in Europe (including Eastern Europe) and Canada, we propose that the importance of low labor
costs and availability of labor as firm-level location factors have a negative effect on the
likelihood of nearshore implementations.
In line with the Uppsala school of management (Johanson & Vahlne, 1977), but also with
the resource-based view and its derivates, we further propose that firm-level offshoring
14
experience and – although not directly measured – advanced capabilities in coordinating offshore
operations may influence location decisions. Similar to our argument for the relevance of
population experience, we argue that firm experience with farshore investments is positively
related to further farshore rather than nearshore investments. This is because, in accordance with
Vahlne & Johanson (1977), we assume for now that offshoring experience with a particular
location reduces psychic distance and also reduces search and coordination costs. This effect
seems to be stronger and more relevant with regard to farshore than it is with regard to nearshore
investments. All these considerations can be summarized in the following propositions:
H4a (Firm-level effect): The more important labor cost savings are in a firm’s offshore
location decision, the less likely is a nearshore implementation.
H4b (Firm-level effect): The more important talent availability is in a firm’s offshore
location decision, the less likely is a nearshore implementation.
H4c (Firm-level effect): The more important geographical proximity is in a firm’s
offshore location decision, the more likely is a nearshore implementation.
H4d (Firm-level effect): The more important cultural proximity is in a firm’s offshore
location decision, the more likely is a nearshore implementation.
H4e (Firm-level effect): The more important language capability is in a firm’s offshore
location decision, the more likely is a nearshore implementation.
H4f (Firm-level effect): The more experience a company has with farshore locations, the
less likely is a nearshore implementation.
In the model developed below we further explore the effect of additional factors we treat as
controls, e.g. firm size, but we also explore the effect of additional firm-level drivers and risks in
15
extended models, e.g. managerial control and client acceptance. Previous research suggests that
these factors matter in location choice (e.g. Lewin & Couto, 2007). However, since they are not
core to the basic model and since they are expected to vary a lot by firm, we do not predict any
effects at this time. However, we will explore and discuss their effects later on and invite future
research to further investigate their impact on nearshore vs. farshore decisions.
Task-level effects: Finally, offshoring research indicates that location preferences very much
depend on characteristics of the task being offshored (e.g. Doh et al., 2009; Hutzschenreuter et
al., 2008). In particular the degree of commoditization of tasks has been stressed as an important
indicator not only for their general ‘offshorability’, i.e. their ‘separability’ from other tasks
(Blinder, 2006), but for the degree of coordination and local firm-specific investment needed to
provide these tasks offshore. High degree of commoditization means that knowledge about a task
is widely spread and diffused across companies and industries, and a potentially large number of
providers and captive centers is able to perform this task. Also, knowledge about performing this
task can be easily ‘transfer’ across locations and organizations. We therefore predict that the
more commoditized tasks are the less important becomes geographical proximity, hence the less
likely is a nearshore investment.
In addition, we recognize that, according to a number of studies (e.g. Dossani & Kenney,
2007; Patibandla & Petersen, 2002; Athreye et al., 2005), India has become the most important
provider of IT and software development services. China, in turn, is specializing primarily in
providing engineering and product development services (A. T. Kearney, 2004; Lewin & Couto,
2007). The concentration of IT, software and product development services in India and China is
also facilitated by the high availability of qualified scientists and engineers in these countries as
16
well as the high degree of digitalization of these services, in particular IT and software, which
eases the transfer and remote coordination of these tasks (e.g. Leonardi & Bailey, 2008). We
therefore predict that offshore investments related to IT, software and product development are
more likely to be made farshore than nearshore. However, we also explore to what extent
administrative functions, e.g. finance and accounting, call centers etc., are more likely to be
nearshored, but, based on previous research we cannot make predictions at this point. From a
theoretical perspective, we focus on the degree of standardization as a predictor for the
(decreasing) relevance of geographical proximity.
H5a (Task-level effect): The more commoditized a task is, the less likely is a nearshore
implementation.
H5b (Task-level effect): The more a task is related to IT, software and product
development, the less likely is a nearshore implementation.
In addition, we seek to explore whether the choice of delivery model (captive or outsourced) is
correlated with nearshore vs. farshore preference. In recent years, specialized external services
providers have become important players in the offshore space, both locally and internationally.
Today, service providers do not just provide IT and business process outsourcing services, but
increasingly develop expertise in product development and analytical services (Couto et al.,
2008). In a decision-making process, the choice of delivery model is often coupled with location
choice. However, previous literature or existing theory does not allow us to make any prediction
about this correlation. We therefore include delivery model merely as a control variable and
discuss our findings later on to promote further research on this aspect.
17
Data and Methodology
We test the effect of the proposed multi-level factors on companies’ decisions to locate business
functions to nearshore rather than farshore locations based on 1,598 offshore implementations by
275 U.S. firms (1,187 implementations) and 88 Dutch firms (411 implementations). The data is
drawn from the ORN database, which is based on an annual survey and which contains
comprehensive data on offshoring strategies, drivers, risks, outcomes and concrete offshore
implementations (both captive and outsourced) of currently 1,322 U.S. and European companies
of all sizes across industries and functions (Lewin & Couto, 2007). 48% of these companies are
actually offshoring, 17% are considering offshoring and 35% are not considering yet. In this
study we solely focus on companies that are currently offshoring. Industries in the ORN database
as well as in our subsample include e.g. manufacturing, software, finance and insurance, and
professional services. Functions offshored include IT, administrative services (e.g. HR, legal,
finance & accounting), call centers, software and product development, marketing and sales, and
procurement. Importantly, offshore implementations reported in the ORN database go back to
the 1980s and early 1990s. However, 99% of implementations reported were made after 1990;
the vast majority was launched in the last five years.
We selected the U.S. and the Dutch sample for this particular study (see Table 1). There
are several reasons this: First, the U.S. and the Netherlands are geographically positioned in
different parts of the world giving them different access to potential nearshore and farshore labor
markets attracting offshore investments. In this respect, the Netherlands represents a typical
Western European country whose companies engage in offshoring. The difference between U.S.
and Dutch (or other Western European) companies in terms of their geographical position makes
them interesting candidates for studying the role of distance in selecting offshore destinations.
18
Second, however, U.S. and Dutch companies are similar in terms of the distribution of functions
being offshored, years of offshoring experience, and delivery models selected (see Table 2-4). In
fact, Dutch companies on average are among the more experienced offshoring firms in Europe.
The similarity between U.S. and Dutch companies in these respects helps control for extraneous
variation (Eisenhardt, 1989). However, the samples do differ in terms of the distribution of firms
by size. The majority of Dutch companies are small and midsize, while the U.S. sample contains
a large amount of larger companies (see Table 5). Not least for this reason, we control for size in
all the regression models below. Third, the U.S. and the Dutch samples are the largest samples in
the ORN database. The high number of offshore implementations allows for a fine-grained
regression analysis and an analysis by subsamples (see in detail below).
-----------------------------
Insert Tables 1-5
-----------------------------
We use a logit model to examine the effect of various multi-level factors on the likelihood of
choosing nearshore vs. farshore locations (dependent variable). Similar approaches have been
taken to study the likelihood of external delivery models vs. captive models (e.g.
Hutzschenreuter et al., 2008) or the likelihood of offshoring product development rather than
other functions (e.g. Lewin et al., 2009). We focus on the likelihood of nearshoring. Nearshore
location is defined as Canada, Mexico and Central America for U.S. companies, and Eastern and
Western Europe for Dutch companies. For both Dutch and U.S. companies, India and China are
the most important farshore locations (see Figure 1). We test the effect of multi-level factors on
the likelihood of nearshoring in different subsamples. The most basic sample includes all U.S.
and Dutch implementations over time. Here we test the effect of multi-level factors (see below)
19
across time and countries of origin. In a second step, we divide the basic sample into two sub-
samples: the U.S. and the Dutch sample, as well as implementations before 2004 and since 2004.
This time split allows for a comparison of multi-level effects on nearshoring vs. farshoring
decisions over time. The year 2004 was chosen mainly for statistical reasons as it divides the
samples roughly into two equally large subsamples. However, it also allows us to separately
analyze dynamics in the most recent years. In a third step, we further divided the samples before
2004 and since 2004 by country which allows us to compare location dynamics by time and
country at the same time.
-----------------------------
Insert Tables 6,7
-----------------------------
Independent variables are listed and explained in Table 6. We test Hypothesis H1 (macro level:
increasing tendency to farshore) mainly by using a 2004 time split dummy (YEARSPLIT) in the
basic sample model. YEARSPLIT is 1 for all implementations since 2004, 0 for implementations
prior to 2004. However, indirectly we contribute to a better understanding of changing
preferences over time by looking at different effects in the subsamples (see below). For testing
H2a and H2b (country level), we, on the one hand, test the effect of national origin of the firm
(i_NL), on the other hand, we use two proxy variables to measure the effect of host country
economic conditions on their likelihood of being selected: COUNTRY_W – the delta between
the log of GDP/capita in the home and host country as a proxy for wage differential, and
COUNTRY_T – the log of the number of students enrolled in tertiary education as a proxy for
size of labor market. Similar measures have been applied by other scholars (e.g. Doh et al.,
2009). Hypothesis H3 (population-level effect) is measured with REL_EXP – the ratio of prior
20
nearshore vs. farshore implementations made across firms of the same national origin. A similar
measure for imitation effects has been used e.g. by Hutzschenreuter et al. (2008).
Hypotheses H4a through H4e (Firm-level location factors) are tested using the variables
LOC_TALENT (importance of talent availability), LOC_LABC (importance of labor costs),
LOC_ PROXG (importance of geographical proximity), LOC_PROXC (importance of cultural
proximity) and LOC_LANG (importance of language capability). All these variables are
standard Likert-scale variables (1 to 5) based on the following ORN survey question: “For each
implementation, please indicate your level of agreement with the following reasons for choosing
this particular country as an offshore location.” (1 – strongly disagree; 5 – strongly agree). In
extended regression models we included other location factors based on this question, e.g.
LOC_MARKET (market proximity), LOC_COLM (collocating with existing manufacturing)
and LOC_CUSTOMER (serving local customers). Another important location factor that is
intended to contribute to the explanatory power of country level effects is LOC_NHOT
(avoidance of hotspots). We included this factor because it indicates to what extent companies
choose a location because it promises to provide better access to qualified, yet lower-cost
personnel relative to other locations. H4f on the effect of firm-level experience is tested using
two simple dummy variables – FARPREV, NEARPREV – which are 1 if the respective
company has made previous farshore or nearshore investment respectively.
Hypotheses H5a and H5b (task-level factors) are tested in the following way: To test the
effect of task commoditization on the relevance of geographical distance (H5a), we use the
variable COMM_TODAY which measures the degree of commoditization of a function, e.g. IT.
This measure is generated from the ORN service provider survey (see also Couto et al., 2008). In
this survey, respondents are asked the following question: “In your opinion, for each class of
21
services that your company provides, how commoditized had this service become? (1 – very low,
5 – very high)”. We took the average response for each function (n= 280 responses) as a proxy
for the degree of commoditization of tasks. To test H5b (effect of software, IT, product
development function), we take SOFTPDIT as a dummy to indicate if the respective function
offshored belongs to the group of IT, software development or product development functions in
the ORN survey. Other surveys, e.g. Doh et al. (2009), use similar measures to characterize the
nature of tasks.
In addition, we use various control variables: EMP_HOME is used as an indicator for
size of the company. It contains the log of the number of employees working for the respective
company domestically. We realize that this size measure can be criticized in that it does not
account for trends towards temporary employment. Also, it does not count employees working at
outsourcers for the company. However, it is an indicator of the ‘resource base’ under
‘ownership’ control of the company. Also, it indicates to what extent a company uses overhead
implying the need for administrative efficiency in order to reduce costs (e.g. Lewin & Couto,
2007). Finally, the control variable i_OUT helps measure the impact of choice of the service
delivery model of preference for nearshore vs. farshore locations. i_OUT is 1 if an
implementation involves either a local, domestic or international service provider, where, in any
case, the task is performed offshore. i_OUT is 0 if the company established a wholly owned
captive unit to perform the task. In addition, we used function dummies (i_IT, i_CC, i_ADMIN
etc.) in certain models to indicate whether a particular function was being offshored.
22
Results
The findings confirm in a nutshell that geographical distance to potential offshore labor markets
matters in the selection of offshore locations (Tables 8-10). Dutch companies, in accordance with
Hypothesis H2a, are much more likely than U.S. companies to select nearshore locations close to
their home country (Model 1, 2, 3), because of their proximity to a larger nearshore labor market.
Availability of qualified personnel is in fact partly positively related to nearshore investments in
the Dutch sample (Model 14). However, the likelihood to offshore to remote locations increases
over time (Model 1), although only some models as well as the descriptive statistic of location
distribution confirm the significance of this trend both for U.S. and Dutch companies. In other
words, the macro-level Hypotheses H1 (tendency towards farshore) can be partially confirmed.
Country-level factors are tested in the extended Models 10-15. Results suggest that, as predicted,
low labor costs and availability of qualified personnel positively influence the preference for
farshore rather than nearshore locations. Hypotheses H2b and H2c can be confirmed. By
contrast, population tendency towards farshore (Hypotheses H3) does not positively affect the
preference for farshore decisions. In other words, population behavior does not have a significant
affect on firms’ location preferences. Hypotheses H3 cannot be confirmed.
-----------------------------
Insert Tables 8-11
-----------------------------
Rather, findings indicate that the firm-level importance of location factors as well as prior
farshoring experience have a significant influence on location decisions (Hypotheses H4a
through H4f). Most striking is the trade-off between the importance of availability of talent and
labor costs on the one hand and the importance of geographical proximity on the other hand (see
23
most models). Findings confirm that the more important availability of talent and low labor costs
are in an offshoring decision, the less likely are nearshore implementations (H4a, H4b). On the
other hand, the more important geographical proximity is, ceteris paribus, the more likely is a
nearshore implementation (H4c). Interestingly, cultural proximity and language capability do not
seem to influence nearshore decisions. Hypotheses H4d and H4e cannot be confirmed. On the
contrary, after 2003, companies, in particular from the U.S., even associate language capability
negatively with nearshore locations (Model 9). In other words, lack of language capability –
beside other reasons – can be even listed as a reason why many U.S. companies do not choose
nearshore locations, in particular Mexico and Central America. Cultural proximity only initially
seems to have supported nearshore investments – in the Dutch sample (Model 6). Later on, it did
not play a significant role.
Findings suggest however that firm-level experience with farshore destinations matters in
location decisions (H4f). Across companies, but especially in the Dutch sample, prior farshore
investments positively influence subsequent farshore investments. On the other hand, previous
nearshore investments do not seem to matter a lot. In other words, inexperienced companies are
more likely to invest nearshore than farshore. This finding supports the Uppsala model of
internationalization (Johanson & Vahlne, 1977), according companies, on the one hand, prefer to
invest in locations characterized by low psychic distance, on the other hand, learn how to reduce
psychic distance to further ‘remote’ locations over time. Interesting however are parallel changes
in the firm-level importance of other location factors. Findings suggest that for all companies,
but in particular for Dutch companies (see also Table 11), low labor costs has become
increasingly important as a location factor, which increases the likelihood of farshore
implementations. For U.S. companies, by contrast, reducing labor costs has always been an
24
important offshoring driver (Table 10, 11). Dutch companies, however, also regard availability
of labor as increasingly important - supporting farshoring (Model 7). Factors like proximity to
markets or supporting local customers used to be important for Dutch companies, but are not
anymore (see Table 11). Yet, these factors do not seem to influence nearshore vs. farshore
decisions significantly (see Model 10-12). Interestingly, for U.S. companies, avoiding hotspots
has become an important criterion for choosing nearshore locations (Model 9). For Dutch firms
this seems to be the opposite (Model 7), although the effect is not significant.
Task-level factors (Hypotheses H5a, H5b) play only a minor role in deciding about
nearshoring vs. farshoring, one exception being IT and procurement in favor of farshoring.
Overall, H5a and H5b cannot be confirmed. Quite interesting, however, is the effect of the
control variables firm size and service delivery model. In particular in the U.S. sample, larger
firms show an increasing tendency to nearshore (Model 3, 5, 9). At the same time, external
delivery models are positively related to nearshoring, in particular in the U.S. sample in more
recent years (Model 3, 9, 15). In the following we discuss some of these findings in greater
detail, paving the way for a range of potential follow up studies. We also list a number of
limitations that need to be addressed in the future.
Discussion
Empirical findings of this study suggest a co-evolutionary dynamic of offshoring location
choices. When U.S and Dutch companies started relocating business functions and processes in
the early 1990s, they faced different initial conditions in their location decisions resulting from
their geographical position. Dutch companies, on the one hand, had the option, in particular after
1990, to tap into a rather large nearshore labor market – first mainly Western Europe, then
25
Eastern Europe, thereby reconciling the benefits of geographical proximity with the opportunity
to hire lower cost talent. U.S. companies, on the other hand, were incented to explore farshore
locations, in particular India, right away because the size of the nearshore labor market
(compared to Western and Eastern Europe) was rather small. Also, most U.S. companies have
been strongly driven by the opportunity to lower labor costs from the very beginning, reinforcing
their orientation towards farshore locations, despite their geographical distance.
The U.S. experience in India arguably sparked an increasing interest in the media and in
consulting, promoting an offshoring agenda that until today stimulates a wide range of debates
and studies. Offshoring has become associated with the ‘global race for talent’ (Manning et al.,
2008) and the opportunity to increase efficiencies globally. This discursive development (see
also Metters & Verma, 2008) might have contributed to the significant increase of ‘awareness’ of
availability of talent and labor cost advantages as offshoring location factors among Dutch
companies. Moreover, Dutch companies are likely to have experienced tightening labor markets
in Eastern Europe. Case studies suggest that many European companies did not think ‘long-term’
when investing in Eastern European locations. Driven by their immediate opportunity to hire
lower cost people nearshore, they in fact contributed to increasing wage inflation and
competition for talent and ‘used up’ the nearshore labor market. Only then alternative options,
e.g. India, came to managerial attention. Another dynamic might be the enlargement of the EU in
2004. Many companies see this affecting opportunities to benefit from labor arbitrage effects.
Consequently, more Dutch firms are now offshoring to India and other offshore locations.
U.S. firms, however, seem to spread out to alternative options. At least two factors seem to play
a role here: As more European companies source talent and external expertise in India, problems
of wage inflation and employee turnover accelerate (e.g. Mayer-Ahuja & Feuerstein, 2007). This
26
might be one reason why more U.S. companies now consider nearshoring – not only because of
geographical proximity, but as an option to avoid hotspot locations. At the same time, however,
specialized, India- and China-based service providers have emerged over time who are
increasingly internationalizing their operations (Couto et al., 2008; Manning et al., 2008). In
order to better serve their longer term U.S. clients, many service providers have by now set up
operations close to the customer, either in the U.S. directly or in nearshore destinations. This has
arguably promoted recent decisions by U.S. companies to invest nearshore. ORN research
suggests that in particular large companies make increasing use of international – rather than
local – service providers. Plus, in particular Mexico and Central America are dominated today by
full-service providers, such as Accenture, IBM and Wipro.
In more abstract terms, findings suggest that sourcing strategies and offshoring location
choices by U.S. and Western European companies are systemically related to each other within a
fast changing global competitive environment. The global search and competition for (lower-
cost) talent is shaped by strategic responses of MNCs – as well as other players, e.g. external
service providers – to changing competitive conditions as well as unintended consequences of
actions at multiple systemic levels. Established theories of internationalization, e.g. the OLI
framework (Dunning, 1981), the Uppsala Model (Johanson & Vahlne, 1977), but also Porter’s
work (Porter, 1990), seem to have underestimated these dynamics. While Porter for example
would argue that MNCs can develop a competitive advantage based on their domestic experience
or – in this context – because of their initial geographical proximity (or distance) to offshore
labor markets, this argumentation falls short of recognizing the ongoing parallel, intersecting
trajectories at the firm, industry and national economy level (see e.g. Volberda & Lewin, 2003;
Lewin & Volberda, 1999). We propose, based on our findings, that future research on offshoring
27
(location) strategies should be more time-sensitive, e.g. by applying concepts like ‘window of
opportunity’ (see e.g. Tyre & Orlikowski, 1994).
Future studies also need to overcome limitations of this study. We focused solely on
companies from the U.S. and Netherlands. Although they are certainly important contributors to
the global offshoring dynamic, other firms might show other strategic patterns. For example,
more than U.S. and Dutch companies, UK and Spanish companies are expected to rely on
historical colonial relationships, e.g. with India and Latin America respectively. Also, we
abstracted away from the fact that in particular companies from the U.S. might differ in terms of
their headquarter proximity to potential offshore locations. Companies at the East coast, for
example, might regard Ireland as a nearshore destination, while Californian companies might be
more oriented towards China and Mexico from the very beginning. Also, as mentioned earlier,
company size might make a significant difference in location choice patterns. Since many U.S.
companies are larger than most Dutch companies, this might influence nearshore vs. farshore
preferences. Finally, one current limitation of ORN data is the lack of differentiation between
‘contact location’ and ‘task location’. In particular the rise of international service providers is
likely to change location choice patterns significantly which, so far, is only partly reflected in
ORN data.
In addition, our findings may stimulate various follow-up research questions explorating
the dynamic nature of internalization in general and offshoring location choices in particular.
Possible follow-up research questions could include to what extent companies observe
competitive responses, and how they, in turn, respond to these observations? In other words,
examining a moderating effect of absorptive capacity on offshoring location choices could add to
the explanatory power of the model. Other interesting research venues might include the role
28
nation states play. To what extent for example are domestic and nearshore policy-makers aware
of these dynamics and how do they respond? And how does migration of talent (see e.g. OECD,
2008) affect location choice patterns? How is talent different, in this respect, from other ‘location
factors’? We suggest, based on our exploratory study, that a co-evolutionary multi-level
framework should be used to better understand these dynamic phenomena.
References
A.T.Kearney. 2004. Making offshore decisions: A.T.Kearney's 2004 offshore location
attractiveness index, Chicago et al.
Athreye SS. 2005. The Indian software industry and its evolving service capability. Industrial
and Corporate Change 14: 393-418
Blinder AS. 2006. Offshoring: the next industrial revolution? Foreign Affairs 85: 113-28.
Bresnahan T, Gambardella A, Saxenian A. 2001. 'Old economy' inputs for 'new economy'
outcomes: cluster formation in the new Silicon Valleys. Industrial and Corporate Change
10: 835-60.
Bunyaratavej K, Hahn ED, Doh JP. 2007. International offshoring of services: a parity study.
Journal of International Management 13: 7-21.
Child J. 1972. Organizational structure, environment and performance: The role of strategic
choice. Sociology 6: 1-22
Couto V, Mani M, Sehgal V, Lewin AY, Manning S, Russell JW. 2008. Offshoring 2.0:
Contracting knowledge and innovation to expand global capabilities, 2007 Service
Provider Survey Report by Booz & Co. & Duke University.
Doh JP. 2005. Offshore outsourcing: implications for international business and strategic
management theory and practice. Journal of Management Studies 42: 695-704
Doh J, Bunyaratavej K, Hahn ED. 2009. Separable but not equal: the location determinants of
discrete services offshoring activities. Journal of International Business Studies
Forthcoming.
Dossani R, Kenney M. 2007. The next wave of globalization: relocating service provision to
India. World Development 35: 772-91.
29
Eisenhardt KM. 1989. Building theories from case study research. Academy of Management
Review 14: 532-50
Dunning, JH 1980. Toward an Eclectic Theory of International Production: Some Empirical
Tests. Journal of International Business Studies 11: 9-31.
Freeman R. 2006. Does globalization of the scientific/engineering workforce threaten U.S.
economic leadership? NBER Innovation Policy & the Economy Working Paper.
Gupta A, Seshasai S, Mukherji S, Ganguly A. 2007. Offshoring: The transition from economic
drivers toward strategic global partnership and 24-hour knowledge factory. Journal of
Economic Commerce in Organizations 5: 1-23.
Hitt MA, Beamish PW, Jackson SE, Mathieu JE. 2007. Building theoretical and empirical
bridges across levels: multilevel research in management. Academy of Management
Journal 50: 1385-99.
Hofstede G. 1980. Culture's Consequences: International Differences in Work-related Values.
Beverly Hills: Sage Publications.
Huang Y, Khanna T. 2003. Can India overtake China? Foreign Policy 137 (July/August): 74-81.
Hutzschenreuter, T, Lewin, AY, Dresel, S. 2008. Governance mode for offshoring of support
functions: Comparison of US and German companies. Paper presented at the 2008
Academy of International Business Conference.
Johanson J, Vahlne J-E. 1977. The internationalization process of the firm - a model of
knowledge development and increasing foreign market commitments. Journal of
International Business Studies 8: 23-32.
Kenney M, Massini S, Murtha TP. 2009. Offshoring administrative and technical work: do we
need a new theory for this? Journal of International Business Studies Forthcoming.
Kogut B, Singh H. 1988. The effect of national culture on the choice of entry mode. Journal of
International Business Studies 19: 411-32.
Leonardi PM, Bailey DE. 2008. Transformational technologies and the creation of new work
practices: Making implicit knowledge explicit in task-based offshoring. MIS Quarterly
32: 411-36.
Levy D. 2005. Offshoring in the new political economy. Journal of Management Studies 42:
685-693.
30
Lewin AY, Couto V. 2007. Next Generation Offshoring: The Globalization of Innovation, Duke
University CIBER/Booz Allen Hamilton Report, Durham.
Lewin AY, Massini S, Peeters C. 2009. Why are companies offshoring innovation? The
emerging global race for talent. Journal of International Business Studies Forthcoming.
Lewin AY, Volberda HW. 1999. Prolegomena on coevolution: A framework for research on
strategy and new organizational forms. Organization Science 10: 519-34.
Manning S. 2008. Customizing clusters: On the role of multinational corporations in the
formation of science and engineering clusters in emerging economies. Economic
Development Quarterly 22: 316-23.
Manning S, Massini S, Lewin AY. 2008. A dynamic perspective on next-generation offshoring:
the global sourcing of science & engineering skills. Academy of Management
Perspectives 22: 35-54.
Mayer-Ahuja N, Feuerstein P. 2007. IT-labour goes offshore: regulating and managing attrition
in Bangalore. SOFI Working Paper.
Metters R, Verma R. 2008. History of offshoring knowledge services. Journal of Operations
Management 26: 141-7
O'Leary MB, Cummings NJ. 2007. The spatial, temporal, and configurational characteristics of
geographic dispersion in teams. MIS Quarterly 31: 433-52.
OECD. 2008. The Global Competition for Talent: Mobility of the Highly Skilled. Paris: OECD
Report.
Oliver C. 1991. Strategic responses to institutional processes. Academy of Management Review
16: 145-79.
Patibandla M, Petersen B. 2002. Role of transnational corporations in the evolution of a high-
tech industry: the case of India’s software industry. World Development 30: 1561-77
Porter ME. 1990. The competitive advantage of nations. New York: Free Press.
Rugman AM, Verbeke A. 2004. A perspective on regional and global strategies of multinational
enterprises. Journal of International Business Studies 35: 3-18
Scott WR. 2001. Institutions and organizations. Thousand Oaks: Sage.
Tyre MJ, Orlikowski WJ. 1994. Windows of opportunity: temporal patterns of technological
adaptation in organizations. Organization Science 5: 98-118
31
Shenkar O, Luo Y, Yeheskel O. 2008. From "distance" to "friction": substituting metaphors and
redirecting intercultural research. Academy of Management Review 33: 905-23.
Volberda HW, Lewin AY. 2003. Co-evolutionary dynamics within and between firms: from
evolution to co-evolution. Journal of Management Studies 40: 2111-36
Xu D, Shenkar O. 2002. Institutional distance and the multinational enterprise. Academy of
Management Review 27: 608-18.
32
Tables and Figures Figure 1: Distribution of location choices by national origin of firm
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
BE DE NL ESP UK USA
( p )
Source: Duke University/Archstone Consulting Offshoring Research Network 2005 Survey and Duke University/Booz Allen Hamilton Offshoring Research Network 2006 Survey and Duke University/The Conference Board Offshoring Research Network 2007/8 Survey
Per
cent
of t
otal
num
ber
of im
plem
enta
tions
N= 193 190
OtherOther AsiaPhilippines
ChinaIndia
Latin AmericaMexico
CanadaEastern EuropeWestern Europe
411 174 103 1187
Figure 2: Distribution of offshore implementations in India by national origin of firm*
11
0%
10%
20%
30%
40%
50%
60%
Pre-2002 2002-2004 2005-2007
SpainBelgiumGermanyScandUKNLUS
Source: Duke University/Archstone Consulting Offshoring Research Network 2005 Survey and Duke University/Booz Allen Hamilton Offshoring Research Network 2006 Survey and Duke University/The Conference Board Offshoring Research Network 2007/8 Survey
Perc
ent o
f tot
al n
umbe
r of
impl
emen
tatio
ns
Year in which offshore implementation was made
*n=776 implementations in ORN database
33
Table 1: Offshoring Status of U.S. and Dutch Companies in ORN Sample
No. of firms
No. of implementati
ons No. of firms
No. of implementa
tions No. of firms
No. of implementatio
nsCurrently offshoring 301 1,271 102 453 403 1,724Considering offshoring 53 182 54 108 107 290Not considering offshoring 73 73 208 208 281 281
Total 427 1,526 364 769 791 2,295
US Netherlands Total
Table 2: Distribution of Functional Implementations in U.S. and Dutch Sample
US Netherlands TotalProduct Development 22% 20% 21%Software Development 20% 16% 18%Administrative Services 17% 13% 15%Information Technology 16% 14% 16%Call Centers 14% 14% 14%Marketing & Sales 5% 13% 8%Procurement 5% 10% 7%Total 100% 100% 100% Table 3: Distribution of Service Delivery Models in U.S. and Dutch Sample
US Netherlands TotalCaptive 46% 58% 49%Outsourced 54% 42% 51%Total 100% 100% 100% Table 4: Distribution of Location Choices Over Time in U.S. and Dutch Sample*
Total Nearshore % Total Nearshore % Total Nearshore %US 373 45 12% 461 38 8% 1,187 135 11%Netherlands 124 48 39% 119 35 29% 411 130 32%Total 497 93 19% 580 73 13% 1,598 265 17%
Total SamplePre-2004 2004-present
*Some firms excluded because launch year data not available
34
Table 5: Distribution of Firms by Size in U.S. and Dutch Sample*
Small (<500
employees)
Midsize (500-
20,000 employe
es)
Large (>20,000
employees) TotalNo. of firms % of total
No. of firms % of total No. of firms % of total
No. of firms % of total
US 69 28% 90 36% 88 36% 247 100%Netherlands 49 67% 24 33% 0 0% 73 100%Total 118 37% 114 36% 88 28% 320 100% *Some firms excluded because size data not available Table 6: List of Variables
Dependent variable Explanation / Hypothesis Format NEARSHORE Nearshore implementation Dutch: 1 for Western or
Eastern Europe; 0 otherwise U.S.: 1 for Canada, Mexico or Central America; 0 otherwise
Independent variables YEARSPLIT (After2003) H1: Implementation after 2003
(50% split ; sample separation by time)
1;0
i_NL H2a: Dutch company (also used for sample split)
1;0
COUNTRY_T H2b: Number of students in tertiary education system
Log(number)
COUNTRY_W H2b: Wage differential measured by log(GDP/Capita home country) – log(GDP/Capita host country)
Number
REL_EXP H3: Population location choice trends (nearshore/farshore)
No. of past farshore / nearshore implementations
LOC_TALENT, LOC_LABC, LOC_PROXG, LOC_LANG, LOC_PROXC
H4: Core location factors (e.g. importance of geographical prox., cultural prox., language comp., labor costs, talent)
5-Point Likert scale (1… not important at all; 5… very important)
LOC_MARKET, LOC_COLMAN, LOC_CUSTOMER, LOC_NHOTSPOT
H4: Additional location factors (e.g. importance of market prox., co-location with manufacturing, serving local customers, avoiding hotspots)
5-Point Likert scale (1… not important at all; 5… very important)
FARPREV, NEARPREV H4f: Firm-level experience with prior farshore, nearshore investments
1;0
COMM_TODAY (COMM_FUTURE)
H5a: Level of commoditization of task (today and in the future)
Based on 5-Point Likert scale (1…very low; 5… very high)
i_SOFTPDIT H5b: IT, Software or Product development function
1;0
i_IT, i_CC, i_ADMIN, i_PROC, i_MS
Control: Function Dummies (IT, Administrative Services, Procurement, Marketing & Sales)
1;0
EMP_HOME Control: Number of employees domestically (size proxy)
Log (number)
i_OUT Control: Dummy for outsourced (rather than captive) operation (service delivery model)
1;0
35
Table 7: Correlations between independent variables
Variable Obs Mean Std. Dev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181 loc_talent 1169 3.69 1.03 12 loc_labcost 1194 4.03 1.07 0.17* 13 loc_proxgeog 1147 2.89 1.33 0.1* -0.06 14 loc_lang 1389 3.13 1.32 0.35* 0.13* 0.37* 15 loc_proxcult 1156 2.99 1.17 0.22* -0.02 0.65* 0.53* 16 loc_nhotspot 1140 2.76 1.08 0.3* 0.12* 0.41* 0.35* 0.41* 17 loc_market 1134 3.01 1.36 0.15* -0.07 0.39* 0.24* 0.35* 0.39* 18 loc_colman 1363 2.70 1.43 0.07 -0.01 0.48* 0.26* 0.35* 0.28* 0.45* 19 loc_customer 1134 2.91 1.36 0.15* -0.1* 0.45* 0.31* 0.39* 0.36* 0.79* 0.55* 1
10 loc_provloc 1372 2.87 1.31 0.44* -0.03* 0.25* 0.29* 0.33* 0.34* 0.26* 0.26* 0.31* 111 i_Out 1399 0.54 0.50 0.04 0.11* -0.11* 0 -0.09* -0.01 -0.24* -0.14* -0.19* 0.23* 112 emp_home 1767 7.29 3.17 0.13* 0.13* 0.12* 0.12* 0.11* 0.06 -0.05 0.04 -0.02 0.07 -0.06 113 country_w 1802 2.60 1.42 0.12* 0.44* -0.28* 0.07 -0.25* -0.06 -0.21* -0.09* -0.24* -0.04 0.1* 0.06 114 country_t 1775 15.20 1.51 0.19* 0.23* -0.21* -0.02 -0.21* -0.09* -0.1* -0.05 -0.12* -0.01 0.04 0.06 0.61* 115 comm_today 1819 3.11 0.46 0.03 0.1* 0.14* 0.14* 0.18* 0.14* 0 0 0.03 0.06 0.1* 0.21* -0.02 -0.15* 116 comm_future 1819 3.53 0.45 0.02 0.09* 0.15* 0.16* 0.19* 0.14* 0.01 0.02 0.04 0.08* 0.11* 0.22* -0.02 -0.15* 0.97* 117 farprev 1081 0.51 0.50 -0.08 0.02 0.01 0.02 -0.04 0.02 0.01 0 0 0 -0.01 0.21* -0.02 -0.04 0 0.01 118 nearprev 1081 0.18 0.38 -0.13* -0.18* 0.08 -0.01 0.01 0.03 0.13* 0.05 0.12* 0.01 -0.01 -0.06 -0.24* -0.23* 0.05 0.06 0.26* 1
Significance level: * <1%
36
Table 8: Regression models (Dependent variable: Likelihood of Nearshoring)
1 2 3 4 5 6 7 8 9Independent Variables Coded Name Total Total Total NL US NL NL US US
Pre-2004 2004+ Pre-2004 2004+ Pre-2004 2004+# Obs 584 263.00 321.00 143.00 441.00 70.00 51.00 177.00 247.00LR Chi-2 215.23 102.58 128.26 69.31 93.20 41.95 30.74 45.23 56.89Prob > Chi2 0.0000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00Log Likelihood -169.009 -91.151 -65.970 -62.916 -98.729 -26.513 -19.181 -47.631 -36.023R-Squared 0.3890 0.3601 0.4929 0.3552 0.3207 0.4417 0.4449 0.3220 0.4412
Constant -3.550*** (1.238) -4.678*** (1.816) -6.250*** (2.239) 1.977 (1.664) -6.190*** (1.964) 0.121 (2.834) 2.830 (2.736) -6.517** (2.780) -8.301*** (3.142)
SubsamplingSurvey Country i_NL 3.118*** (0.449) 3.259*** (0.704) 4.287*** (0.871) n/a n/a n/a n/a n/a n/aAfter 2003 yearsplit03 -0.478* (0.291) n/a n/a -0.453 (0.510) -0.596+ (0.387) n/a n/a n/a n/a
Location FactorsTalent Availability loc_talent -0.587*** (0.170) -0.564** (0.239) -0.512** (0.260) -0.514* (0.278) -0.719*** (0.245) -0.629 (0.478) -0.943+ (0.608) -1.332*** (0.384) -0.238 (0.390)Low Labor Cost loc_labc -0.449*** (0.132) -0.198 (0.205) -0.808*** (0.224) -0.459** (0.214) -0.327+ (0.203) -0.424 (0.404) -1.086** (0.432) -0.003 (0.321) -0.696* (0.364)Geo Proximity loc_proxg 1.133*** (0.166) 1.185*** (0.245) 1.385*** (0.296) 1.167*** (0.257) 1.286*** (0.298) 1.328*** (0.428) 0.874* (0.475) 1.436*** (0.487) 1.647*** (0.526)MatcLanguage loc_lang 0.049 (0.159) 0.156 (0.228) -0.414* (0.251) -0.137 (0.263) 0.191 (0.229) -0.687 (0.511) 0.547 (0.539) 0.733* (0.387) -0.724* (0.382)Cultural Proximity loc_proxc 0.040 (0.174) 0.234 (0.238) -0.136 (0.309) -0.097 (0.276) 0.158 (0.278) 0.719* (0.433) 0.570 (0.678) 0.280 (0.418) -0.083 (0.466)Avoiding Hotspots loc_nhot 0.036 (0.162) 0.038 (0.229) 0.238 (0.286) -0.297 (0.303) 0.180 (0.218) 0.391 (0.654) -0.593 (0.539) -0.093 (0.304) 0.729+ (0.464)
Company FactorsService Delivery Model i_Out 0.366 (0.313) 0.004 (0.426) 0.896+ (0.567) -0.200 (0.548) 0.466 (0.435) 0.432 (1.039) 1.297 (1.148) 0.261 (0.566) 1.460+ (0.968)Size of Firm emp_home 0.132** (0.064) 0.080 (0.084) 0.354*** (0.132) 0.027 (0.157) 0.175** (0.077) -0.466 (0.356) -0.059 (0.332) 0.063 (0.092) 0.335** (0.154)
Function VariablesIT i_IT -0.677 (0.601) -1.401* (0.828) 0.620 (1.008) -1.763+ (1.121) -0.255 (0.857)Call Center i_CC -0.070 (0.586) -0.649 (0.792) 1.410 (1.018) -0.238 (0.873) 0.051 (0.884)Product Dev or Softwarei_PDSoft -0.121 (0.550) -0.699 (0.744) 1.377 (0.981) -0.044 (0.724) -0.067 (0.888)Administrative i_Admin -0.454 (0.657) -1.278 (0.922) 1.516 (1.154) 1.264 (1.499) -0.551 (0.904)Procurement i_Proc -1.046 (0.735) -1.277 (0.936) -0.012 (1.369) -2.195* (1.171) -0.142 (1.040)Marketing Services i_MS -0.252 (0.677) -1.503* (0.885) 2.744** (1.303) -0.512 (0.911) -0.705 (1.366)Software or PD or IT SoftPDIT 0.690 (0.963) -1.542 (1.430) -0.166 (0.585) -0.422 (0.682)
Standard Errors in brackets. Significance levels: *** <1%, ** < 5%, * < 10%, +<15%.
37
Table 9: Extended regression models (Dependent variable: Likelihood of Nearshoring)
10 11 12 13 14 15Independent Variables Coded Name Total NL US Total NL US
# Obs 496.00 103.00 393.00 526.00 105.00 421.00LR Chi-2 274.17 96.47 137.21 272.07 115.02 130.33Prob > Chi2 0.00 0.00 0.00 0.00 0.00 0.00Log Likelihood -88.526 -22.337 -58.491818 -101.494 -14.465 -71.477R-Squared 0.6076 0.6835 0.5398 0.5727 0.7990 0.4769
Constant 9.76*** (3.22) 23.245*** (8.97) 5.208 (3.962) -4.412 (5.381) -36.352** (17.636) 4.937 (8.293)
SubsamplingSurvey Country i_NL 2.75*** (0.63) n/a n/a 3.355*** (0.623) n/a n/aAfter 2003 yearsplit03 0.28 (0.44) 1.587 (1.36) 0.024 (0.545) 0.285 (0.409) 3.837** (1.874) -0.169 (0.498)
Location FactorsTalent Availability loc_talent -0.76*** (0.28) -1.097 (0.78) -0.728* (0.380) -0.428* (0.231) -2.385 *(1.302) -0.226 (0.294)Low Labor Cost loc_labc -0.02 (0.21) -0.016 (0.50) 0.121 (0.281) -0.355** (0.173) -1.958** (0.892) -0.235 (0.242)Geo Proximity loc_proxg 1.37*** (0.25) 2.220*** (0.85) 1.491*** (0.380) 1.325*** (0.222) 4.063*** (1.499) 1.402*** (0.329)MatcLanguage loc_lang -0.14 (0.24) -0.724 (0.70) -0.021 (0.323) 0.082 (0.214) -2.085* (1.180) 0.216 (0.278)Cultural Proximity loc_proxc -0.09 (0.28) 0.283 (0.61) -0.112 (0.384) -0.135 (0.238) -0.745 (0.985) -0.044 (0.334)Avoiding Hotspots loc_nhot -0.04 (0.24) -0.822 (0.86) 0.056 (0.297) -0.048 (0.215) 1.922 (1.462) -0.013 (0.255)Market Proximity loc_market 0.13 (0.25) 0.385 (0.73) -0.027 (0.339)Collocating with existing Maloc_colm -0.16 (0.18) -0.575 (0.52) -0.090 (0.225)Customers loc_customer -0.18 (0.24) 0.305 (0.72) -0.183 (0.357)Best Provider Location loc_provloc 0.29 (0.22) -0.156 (0.54) 0.471+ (0.297)
Company FactorsService Delivery Model i_Out 0.41 (0.48) 2.067 (1.54) 0.093 (0.096) 0.767* (0.426) 1.455 (1.386) 0.843+ (0.527)Size of Firm emp_home 0.11 (0.09) 0.487 (0.37) 0.372 (0.615) 0.155** (0.074) 0.648 (0.506) 0.144 (0.083)
Market FactorsLocation Wage Level country_w -0.80*** (0.22) -1.534** (0.61) -0.699*** (0.263) -1.043** (0.409) -7.068** (2.893) -0.043 (0.580)Advanced Education Enrollmcountry_t -0.74*** (0.15) -0.948*** (0.34) -0.728*** (0.186) 0.185 (0.447) 4.300** (1.942) -0.877 (0.699)Working Population country_l
ExperienceRelative Exp rel_expPrev Farshore (dummy) farprev -1.333*** (0.459) -3.669** (1.482) -0.679 (0.567)Prev Nearshore (dummy) nearprev 0.572 (0.472) 0.613 (1.563) 0.198 (0.595)
Function VariablesSoftware or PD or IT SoftPDIT -0.09 (0.48) -0.625 (1.61) 0.297 (0.567) 0.340 (0.367) 0.331 (1.703) 0.498 (0.490)
Commoditization of Function (Today) 2.68 (2.16) 10.893+ (7.32) -0.065 (2.698)Commoditization of Function (Future) -3.07 (2.30) -12.668+ (8.03) -0.024 (2.812)
Standard Errors in brackets. Significance levels: *** <1%, ** < 5%, * < 10%, +<15%.
38
Table 10: Strategic Offshoring Drivers of U.S. and Dutch Companies
US Netherlands TotalLabor cost savings 91% 61% 82%Growth strategy 70% 78% 72%Other cost savings 68% 68% 68%Access to qualified personnel 69% 57% 65%Part of a larger global strategy 58% 74% 63%Competitive pressure 60% 52% 58%Increasing speed to market 48% 58% 51%Business process redesign 47% 54% 49%Improved service levels 45% 49% 46%Exploit country-specific advantages 40% 52% 45%Exploit location-specific advantages 38% 51% 43%Domestic shortage of qualified personnel 31% 51% 40%Accepted industry practice 38% 30% 36%Access to new markets 24% 42% 29%Differentiation strategy 27% 30% 28%Enhancing system redundancy 25% 23% 24% Percentage of respondents rating driver as “important” or “very important” (4 or 5 on Likert Scale) Table 11: Location Selection Factors of U.S. and Dutch Companies
Pre-2004
2004-present Total Pre-2004
2004-present Total
Low cost of labor 84% 80% 81% 47% 69% 59%High level of expertise 78% 72% 75% 58% 64% 59%Talent pool available 71% 69% 69% 58% 60% 54%Low costs (besides labor costs) 73% 62% 66% 47% 65% 56%Matches language requirements 62% 55% 57% 31% 33% 31%Access to local market 52% 39% 44% 59% 31% 45%Collocating with existing BP facility offsho 64% 49% 51% 25% 20% 22%Supporting existing customers locally 53% 39% 44% 56% 29% 40%Collocating with existing manufacturing p 48% 38% 43% 30% 26% 27%Geographical proximity 56% 41% 45% 24% 25% 25%Location of the best service provider 36% 45% 40% 40% 41% 37%Cultural proximity 47% 41% 41% 41% 23% 32%Quality of infrastructure 32% 37% 34% 37% 37% 35%Political stability in host country 40% 32% 33% 38% 25% 31%Government incentives 35% 30% 32% 28% 30% 27%Avoiding "hot spots" 35% 29% 30% 15% 18% 16%Other 28% 30% 29% 1% 8% 5%
US Netherlands
Percentage of respondents rating driver as “important” or “very important” (4 or 5 on Likert Scale)