Hello, please answer the below points:
a. Define and distinguish between subcontracting, outsourcing, and offshoring.
b. Discuss the advantages and disadvantages of subcontracting, outsourcing or offshoring jobs.
c. How has technology impacted subcontracting, outsourcing, and offshoring of jobs?
d. Should a worker in today’s global economy have any reasonable expectation of job security? Explain your reasoning.
e. How can unions work with employers to create good jobs for domestic workers?
Provide a detailed post with a minimum of 200 words in APA format. Include the following two sources properly cited and referenced: (a) Holley, W. H., Jennings, K. M., & Wolters, R. S. (2012). The labor relations process (10th Ed.). Mason, OH: South-Western. (b) Productivity Impacts of Offshoring and Outsourcing: A Review by Karsten Bjerring Olsen
Attached: Article b
Statistical Analysis of Science, Technology and Industry
STI Working Paper
Productivity Impacts of
Offshoring and Outsourcing:
Karsten Bjerring Olsen
OECD Directorate for
Science, Technology and Industry (STI)
ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT 2006
Organisation de Coopération et de Développement Economiques
Organisation for Economic Co-operation and Development 06-Mar-2006
English – Or. English
DIRECTORATE FOR SCIENCE, TECHNOLOGY AND INDUSTRY
PRODUCTIVITY IMPACTS OF OFFSHORING AND OUTSOURCING: A REVIEW
STI WORKING PAPER 2006/1
Statistical Analysis of Science, Technology and Industry
Karsten Bjerring Olsen
Document complet disponible sur OLIS dans son format d’origine
Complete document available on OLIS in its original format
English – Or. English
STI Working Paper Series
The Working Paper series of the OECD Directorate for Science, Technology and Industry is designed
to make available to a wider readership selected studies prepared by staff in the Directorate or by outside
consultants working on OECD projects. The papers included in the series cover a broad range of issues, of
both a technical and policy-analytical nature, in the areas of work of the DSTI. The Working Papers are
generally available only in their original language – English or French – with a summary in the other.
Comments on the papers are invited, and should be sent to the Directorate for Science, Technology
and Industry, OECD, 2 rue André-Pascal, 75775 Paris Cedex 16, France.
The opinions expressed in these papers are the sole responsibility of the author(s) and do not
necessarily reflect those of the OECD or of the governments of its member countries.
Copyright OECD, 2006
Applications for permission to reproduce or translate all or part of this material should be made to:
OECD Publications, 2 rue André-Pascal, 75775 Paris, Cedex 16, France.
PRODUCTIVITY IMPACTS OF OFFSHORING AND OUTSOURCING: A REVIEW
Karsten Bjerring Olsen
Despite the attention that offshore outsourcing currently demands in the public media, there is little
empirical evidence on its economic impact. As a consequence of rising fears of job losses associated with
the phenomenon, most existing research on the subject is primarily concerned with addressing related
labour market issues. The impacts on productivity, however, have received only little attention. This paper
surveys the empirical literature on offshore outsourcing and its productivity effects. Due to the small
number of existing studies, the survey also includes research that may serve as indirect evidence of the
phenomenon’s link to productivity, such as its effect on skill upgrading. The most apparent conclusion
drawn from the review is that there appears to be no clear patterns as to how offshore outsourcing affects
productivity, and that much depends on both sector and firm-specific characteristics. There are some
indications, however, that positive productivity effects from foreign material sourcing depends on the
degree to which firms are already globally engaged, but also that such engagements generally could be
close to their optimum level in developed economies. There is little existing research on offshoring of
services, but it appears that its productivity enhancing effects generally are small in manufacturing plants
while being of a somewhat greater magnitude for firms in the services sector.
L’IMPACT DES DÉLOCALISATIONS SUR LA PRODUCTIVITÉ : VUE D’ENSEMBLE
Karsten Bjerring Olsen
Malgré l’intérêt que les délocalisations à l’étranger suscitent dans les médias, on dispose de peu
d’éléments empiriques sur leur impact économique. Le phénomène de délocalisation faisant craindre de
plus en plus des pertes d’emplois, la plupart des études qui y sont consacrées s’attachent essentiellement
aux aspects qui ont trait au marché du travail, l’impact sur la productivité ne retenant guère l’attention. On
commentera dans ce document les recherches empiriques sur les délocalisations à l’étranger et leur impact
en termes de productivité. Vu le petit nombre d’études disponibles, on prendra également en compte les
travaux qui éclairent indirectement le lien avec la productivité, notamment du point de vue de
l’amélioration des qualifications. La conclusion la plus nette qui ressort de ce panorama est la suivante : il
ne se dégage aucun profil clair quant à la façon dont la délocalisation à l’étranger influe sur la productivité,
les caractéristiques du secteur et de l’entreprise jouant à cet égard un grand rôle. Certains éléments
montrent néanmoins que l’impact positif que peut avoir la délocalisation matérielle à l’étranger est
fonction du degré d’implication mondiale de l’entreprise, cette implication pouvant en général être proche
de l’optimum dans les économies développées. Les recherches sont peu nombreuses sur la délocalisation
des services ; il apparaît néanmoins que les gains de productivité dont la délocalisation s’accompagne dans
le secteur manufacturier sont généralement faibles, alors qu’ils sont un peu plus nets dans le secteur des
PRODUCTIVITY IMPACTS OF OFFSHORING AND OUTSOURCING: A REVIEW
Contracting out business activities to foreign providers, or what has come to be called “offshoring”,
has been undertaken for decades. As such, offshoring is not as recent a phenomenon as the impression one
may draw from its current attention in the public media.1
The phenomenon, however, appears to have
entered into a new stage with offshoring of services becoming increasingly important. This change in
structure has been underway for some time, and is generally attributed to the interplay between three
factors: technological advances, economic and competitive pressures to reduce costs and improve
productivity, and institutional developments favouring trade liberalisation.
Advances in technology have reduced transportation costs considerably, but more importantly, the
development and rapid dissemination of information technologies have had enormous economic impacts
through the transformation of work processes, organisational structures as well as on how we
communicate. Large parts of the economy have become digitized which has enabled business activities to
be conducted in entirely new ways, as well as across large distances. This has opened possibilities of trade
in a variety of services that were traditionally non-tradable, and caused a growth potential in offshoring
activities facilitated by the opening of markets at both global and national levels.
Offshoring in services emerged already in the late 1980s and early 1990s with the contracting out of
tasks related to customer services, but has since moved on to a broader range of activities including
engineering, software development and other tasks requiring high-skilled human capital. Due to the
significant size of the services sector in the Western economies, and the increasing broad range of tasks
exhibiting offshoring potential, the number of jobs which could potentially be affected through this channel
As a result, the subject has become increasingly politically charged.
The potential negative impact of offshoring on wages and employment, and the consequential
reactions from the public in the debate, is probably the main reason why most research on the phenomenon
and its economic impacts has focussed on concerns related to labour markets. However, proponents of
offshoring continue to argue that the phenomenon has long term economic benefits, and that it eventually
will increase living standards in OECD countries through positive productivity effects and reductions in
factor costs. Yet, little rigorous research on offshoring and its impacts on productivity or firm performance
1. Illustrations of media focus often refer to an increase in newspaper articles on the subject. While unable to
obtain such figures, a Google search for “outsourcing”, “offshoring” and “offshore outsourcing” at the start
of this paper (1 May 2005), and at its completion (end October, 2005) is illustrative. Hits at the beginning
numbered 13, 1, and 0.5 million for “outsourcing”, “offshoring” and “offshore outsourcing”, respectively
and 75, 3 and 2 million at the completion.
2. A recent study estimated the occupations potentially affected by offshoring (in 2003) to be around 19.2%
of total employment in the EU15, 18.1% in the United States, and 18.6% in Canada. In Korea, the share
was slightly lower, at 13%. It should be noted that these numbers may represent an “upper limit” as not all
jobs in these occupations would necessarily be offshored (OECD, 2004).
has been conducted. It is the purpose of this paper to give an overview of existing research on offshore
outsourcing and its impact on productivity.
The paper consists of seven sections. The following section outlines the issue and discusses the
importance of empirical inquiry as regards the productivity effects of outsourcing. Section three briefly
outlines the theoretical links between outsourcing and productivity, while section four focuses on some of
the empirical methodologies employed in analysing the issue. Section five presents a range of existing
empirical evidence while section six follows with a discussion of a more general nature. Section seven
provides the reader with a summary.
2. Offshoring, outsourcing and offshore outsourcing
There are countless reports on the magnitude and growth of offshoring. The most famous of these
studies is probably an estimate by Forrester.3
While there is no need in the context of this paper to
summarise the different estimates here, the vast number of studies on offshoring themselves appear to
indicate a significant increase in the degree of internationalisation.
Despite the many accounts on the subject, it is difficult to get a clear picture of the impact of offshore
outsourcing on the economy. First of all, as the public debate on the subject has evolved, the terminology
being used has become increasingly blurred. Secondly, ideological considerations have increasingly
entered the debate. And lastly, yet probably most importantly, the lack of hard evidence of the impact of
offshoring has allowed the debate to continue without the merit of a strong information base. Addressing
this lack of empirical evidence is one objective of this paper and of the OECD work on globalisation more
The term “offshoring” is often associated with “outsourcing” but neither implies the other. Whereas
outsourcing refers to the relocation of jobs and processes to external providers regardless of the provider’s
location, offshoring refers to the relocation of jobs and processes to any foreign country without
distinguishing whether the provider is external or affiliated with the firm. Outsourcing may therefore
include job relocations both within and between countries, whereas offshoring refers only to international
relocations. The term offshore outsourcing therefore only covers the relocation of jobs or processes to an
external and internationally located provider.4
3. Forrester, an independent technology research company, estimates that by the year 2015, approximately
3.3 million jobs in the United States would be lost due to offshoring. See Kierkegaard (2003) for a
discussion on this estimate.
4. The key to determining what process is being considered lies in knowing what is internally provided, and
what is externally provided. However, this has become increasingly difficult due to a variety of complex
organisational forms. Mergers, strategic alliances, and public-private partnerships, for instance, have
blurred the border between internal/external, and a number of business strategy terms related to such
enterprise restructuring have only complicated the picture. Moreover, many companies have developed
contracting strategies that are even more complex.
Figure 1. An illustrative matrix of insourcing, outsourcing and offshoring
Within countries Between countries
Domestic outsourcing International outsourcing
Domestic supply International insourcing
Some commentators have criticized the term “offshoring” due to its lack of meaning in especially the
European context, but even in the United States, where “offshoring” prevalently is understood as the
relocation of jobs to countries such as India and China, other terms have emerged, for instance
“nearshoring” in dealing with offshoring from the United States to e.g. Canada and Mexico. As a result,
terms such as “global”, “international” or “cross-border” outsourcing have therefore been suggested.
Nevertheless, for the sake of simplicity this paper will make use of “offshore outsourcing”, “offshoring”
and “outsourcing” as described above. Moreover, it should be noted that the term will refer to both
manufacturing and services inputs, unless otherwise stated.
2.1. Contextualising the phenomenon
To balance the many forecasts of job losses quoted in the public media, analysts of offshoring, be it
economists, consultant companies or international organisations, often relate projected job losses to general
employment changes when evaluating the overall consequences. For example, many studies compare to
US employment turnover, where the annual amount of job destruction alone is estimated to be between
7 and 8 million jobs – a number that dwarfs even the largest forecasts of job-losses linked to offshoring.
Slaughter (2004) draws attention to a less quoted figure, namely the increase in US employment due to
international insourcing from foreign countries which grew from 2.6 million jobs in 1987 to 5.4 million in
2002. Also, contrary to many beliefs, Kirkegaard (2003) notes that between 1999 and 2002 the majority of
job losses in the category “occupations at risk of offshoring” did not happen in services but in the
manufacturing industries. Moreover, these were generally low-wage jobs, with services employment in the
same occupational category increasing.
For the United States it is also commonly noted that the large increase in manufacturing productivity
growth, and not offshoring, has been the main reason for both the decrease in manufacturing employment
and the “jobless recovery” (see i.e. Mann, 2003; Kirkegaard, 2003; Drezner, 2004; Schultze, 2004). Other
explanatory factors with more significance than offshoring are mentioned by other authors including weak
domestic demand and the overvaluation of the dollar (i.e. Baily and Lawrence, 2005).5
explanation suggests that the rate of job creation following the recent recession has been unusually slow
compared with job destruction due to an ongoing structural transformation of the economy (Groshen and
5. For Japan, the increase in overseas employment linked to international outsourcing accelerated in the late
1980s; this coincided with a sharp appreciation of the Yen.
Potter, 2003). Attention has also been drawn to the US net surplus in services trade (Amity and Wei,
2004c), and to a finding that United States imports of business, professional and technical services have
remained constant between 1997 and 2003 (Schultze, 2004). There are countless other contextualising
examples, but the main message seems to be that, fundamentally, offshoring is primarily a trade
phenomenon qualitatively similar to conventional trade (Bhagwati et al., 2004).
Another set of arguments related to offshoring refer to the potential long term benefits of the
phenomenon. For instance, the large amount of relatively high-quality jobs created in less developed
countries due to offshoring may provide these countries with substantial economic benefits. They can build
a stronger economic base, increase domestic consumption, and therefore foster imports from developed
nations. Moreover, offshoring may also enhance living conditions in OECD countries as the increased
productivity and cost reductions stemming from this process will eventually lower product prices and drive
up real wages. In addition, there may be economic benefits from reemployment of those who lose their job
to offshoring, as these people may eventually move into more rewarding occupations, although temporary
adjustment costs of this transition will be incurred (OECD, 2005). The latter argument obviously assumes
the existence of suitable and rewarding reemployment possibilities and is often rationalised by drawing on
3. Theoretical perspectives of outsourcing
To compete in increasingly competitive economic environments, decisions to offshore company
activities are essentially driven by factors related to costs of production, distribution and productivity.
From the perspective of the firm offshoring is therefore seen as a part of its business strategy. For instance,
if offshoring enables a firm to relocate its relatively inefficient production processes to external providers
with cheaper and perhaps more efficient production capabilities, the firm can turn its focus to areas where
it has a comparative advantage and expand output, or engage in new business activities. Offshoring of
tasks with high complexity levels or of core activities, on the other hand, remains less attractive because of
security issues such as the potential lack of control over the processes.6
The theoretical literature on the firm’s decision to produce in-house or outsource through market
contracts is extensive and dates back to Coase (1937) and his theory of the firm. Recently, however, the
attention on the foreign aspects of the phenomenon has grown (see i.e. Antràs and Helpman, 2004;
Antràs et al., 2005, and Grossman and Helpman, 2002). The majority of this work has most commonly
focused on either transaction cost theory, or, in particular, the principal-agent framework.
• Agency theory: According to this theory bounded rationality and self-serving behaviour or
opportunism of a firm’s employees can imply productivity losses. As such, conflicting goals and
interests between the firm and its employees may pose a problem to the firm. To reduce
inefficiencies stemming from this source the firm can outsource its activities to an external
provider and control the output or effort of the provider through an outcome based contract.
• Transaction cost theory: According to this theory, outsourcing is only desirable as long as the
costs of related asset specific investments, contractual incompleteness and search efforts are
lower than the expected cost advantage. In the case of outsourcing it is often linked to
specialization, i.e. to management theories focusing on the firm’s core competences.
Outsourcing could also have productivity enhancing effects at a more aggregate level if offshoring
would lead to the creation of new firms, and the destruction of old ones (Antràs et al., 2005). This process
is often associated with Schumpeter’s theory of creative destruction, and numerous empirical studies have
6. For related empirical evidence see Baker and Hubbard (2003), Masten (1984), and Ono and Stango (2005).
provided support for its positive impacts on productivity (Bartelsman et al., 2003).7
Other relevant theories
have examined the potential for productivity enhancing effects due to knowledge spill-over as well as
firms’ abilities to focus on core competences by outsourcing relatively inefficient activities.
4. Empirical literature and methodology
4.1. Outline of existing research
Given the amount of attention that offshore outsourcing demands in the public media surprisingly
little rigorous empirical research has been done on its economic impacts. The main focus of the studies that
exist deals primarily with labour market issues, which remains true even as research on the subject is
expanding. Direct investigations of outsourcing and its impact on firm productivity are relatively few in
number. Under reasonable assumptions, however, one may draw important indirect linkages between
productivity and skill intensity, and since this subject is a common theme in the empirical literature on
outsourcing and its labour market impacts, some of the results from these studies will also be discussed in
Whether the empirical literature on offshoring concerns productivity or issues related to labour
markets, most studies only cover manufacturing industries. Moreover, many studies focus only on the
relocation of manufacturing processes ignoring the relocation of services. Because of the increasing
importance of services offshoring, especially within the services sector itself, it is unfortunate that insights
regarding productivity effects of the phenomenon primarily have to be drawn from evidence for
manufacturing industries. Most studies are micro oriented and the productivity impacts on manufacturing
establishments are therefore the main focus of analysis. To compensate for a lack of information on
services, a brief overview of different offshoring surveys will be covered. Due to these limitations, the
survey in this paper can give only a fragmented picture of the subject. This, however, reflects the state of
the available evidence, which is still more limited than desirable.
The impact of outsourcing on productivity is commonly analysed by estimating labour productivity
through a production function framework. Other, but less frequently used approaches include the
estimation of total factor productivity (TFP), TFP growth breakdowns and ANOVA analysis.
The standard approach used to study outsourcing’s impact on productivity, especially at the level of
the firm, is to estimate a production function given by
( , ) Yi,t = Ai,tF Ki,t Li,t
where Y refers either to output or value added, A is the technology factor, K is capital, and L is labour. The
subscript i refers to either industry or firm depending on the type of data used, whereas t refers to time. The
functional form is usually assumed to be a Cobb-Douglas function. In this case, dropping the time
subscript, taking logs and subtracting l (where i Li l = ln , etc.) from both sides yields the following basic
expression for the labour productivity level
i i i i i i y l a k l l 1 2 − = + β ( − ) + β
7. Empirical estimates for this factor are also avialable for Portugal (Carreira and Teixeira, 2003), Slovenia,
(Loecker and Konings, 2004), Finland (Maliranta, 2003), Estonia (Masso, Eamets, and Philips, 2004), and
New Zealand (McMillan, 2004).
It is common to let outsourcing take effect through the technology factor of the production function.
This basically means that outsourcing may alter the firm’s underlying production technology by allowing it
to shift the intercept of the log-linear production function. As such, the base regression equation takes the
i i i i i i y − l = β + β k − l + β l + β x + β x + ε 0 1 2 3 4 ( )
where x refers to the measure of outsourcing, and m and s indicate material and services outsourcing
respectively. The remaining production technology factors are picked up by the constant β 0 and the error
term i ε . In cases where the growth rate of labour productivity is estimated (i.e. by first-differencing the
base equation above), the outsourcing variable is often included as a firm-specific effect which implies that
the terms stay as they are. For completeness, the base regression equation then becomes the following
i i i i i i ∆ y − l = + ∆ k − l + ∆l + x + x + u 0 1 2 3 4 ( ) α α ( ) α α α
For industry level estimation, however, the outsourcing variables are most often also included as changes.
The effects of outsourcing on productivity levels and growth are thus determined by the level of
significance of the estimated coefficients on x.
This estimation method is likely to be faced with some constraints. Problems of endogeneity can be
expected since the results might be driven by unobserved covariates that are correlated with both
productivity and outsourcing. For example, plants with a high productivity level may also be more
skill-intensive and for that reason more likely to engage in outsourcing of low-skill intensive processes in
order to focus on core competences. This may leave the regression estimates biased and inconsistent. A
way of reducing such concerns is to use lagged variables of the outsourcing variable or IV estimation,
which is widely used in the literature. Moreover, looking at growth rates by first-differencing is known to
exacerbate potential problems of measurement errors in the data (Griliches and Hausman, 1986).
The analytical framework used for productivity breakdowns is based on a consolidation framework
initiated by Leontief (1967). Services inputs into manufacturing, for example, are reduced into their
constituent elements of labour, capital, and goods inputs. The change in TFP growth in manufacturing can
then be broken down into effects stemming from the purchases of services and from the rate of material
productivity growth in the industry itself (see i.e. ten Raa and Wolff, 2001). If the estimated increase in the
rate of TFP growth computed with the consolidated measure is smaller than the increase in the rate of TFP
growth computed with the direct coefficients, then this could indicate a positive productivity effect of
services outsourcing on the manufacturing industry.
ANOVA analysis, or analysis of variance, is essentially a generalisation of the t-test allowing one to
look for differences in data over multiple groups. In connection with productivity and outsourcing it is used
to look at differences in labour productivity over different levels of outsourcing intensities. This method is
not commonly used in the literature.
Insights into the link between outsourcing and productivity may also be derived from indirect
analysis, i.e. research that does not deliberately seek to establish a relationship between offshore
outsourcing and productivity, but that nevertheless may give insights into how these variables could be
linked. For example, analysis of the impact of firm characteristics on the decision to outsource could be
valuable. Kimura (2002) and Tomiura (2004), for example, both estimate the determinants of outsourcing
decisions and include productivity as one of their explanatory variables. Their analysis framework,
however, indicates that the causality between outsourcing and productivity is not as clear cut as is often
argued, and that investigating the link between outsourcing and productivity is not without problems.
Other useful insights can be obtained by addressing how outsourcing affects the skill intensity in
industries and establishments. Such studies exist for the United States (Feenstra and Hanson, 1999), the
United Kingdom (Hijzen, 2003; and Hijzen, Görg and Hine, 2003), the European Union (Egger and Egger,
2001a), Germany and Italy (Helg and Tajoli, 2004), France (Strauss-Kahn, 2002), and Japan (Head and
Ries, 2002). Skill intensity is often measured as the costs of high-skilled labour relative to the total wage
bill. Thus, assuming that the marginal product of high-skilled workers is higher than that of
low-skilled workers, a positive effect of outsourcing on a firm’s skill-intensity would, ceteris paribus, also
indicate a positive productivity change.
4.3. Measures of outsourcing
A common measure of offshore outsourcing used in the literature is derived, to a greater or lesser
extent, from that provided by Feenstra and Hanson (1996a, 1996b, 1999). They estimate offshore
outsourcing as the share of imported intermediate inputs over total costs, which for each industry i can be
⎛ = ∑
where is input purchases of good j by industry i, is total non-energy input used by industry i, is
import of good j, and is the consumption of good j. Based on this measure they calculate a “narrow”
measure of outsourcing by restricting attention to those inputs that are purchased from the same industry as
that in which the good is being produced (using 2-digit SIC industry codes). They also employ a second
measure which they call “differential outsourcing”. This is simply calculated as the difference between
their total outsourcing estimate and narrow outsourcing.
j Xi Yi M j
While especially the narrow measure defined by Feenstra and Hanson is widely used, and is in line
with the WTO mode 1 definition of international outsourcing, there is no consensus that it is the most
appropriate measure. Girma and Görg (2004), for example, argue that it is too wide, especially for analyses
at the establishment level. Instead they prefer a measure suggested by Abraham and Taylor (1996) which
includes only the contracting out of machine maintenance services, engineering and drafting services,
accounting services, computer services, and janitorial services. Egger and Egger (2001a) and Helg and
Tajoli (2004) also use a narrower measure restricting offshore outsourcing to outward processing. This
measure includes only the intermediate exports for processing that are re-imported. Others like Görg et al.
(2004) and Criscuolo and Leaver (2005) have more direct data on intermediate inputs, including e.g. raw
materials and components, and services inputs as well as the proportion of these sourced abroad. A
discussion on the measurement issues associated with offshoring is given in OECD (2005a), albeit with a
focus on related labour relocations.
5. Empirical evidence
5.1. Productivity effects at the industry level
As early as the 1960s, Baumol (1967) argued that one of the major reasons for the growth of the
services sector could be linked to services outsourcing in the manufacturing industries (see also Fixler and
Siegel, 1999; ten Raa and Wolff, 2001). The fact that manufacturing industries to an increasing extent
could outsource their less efficient service activities, and focus on their core competences, led to significant
productivity gains in the industry, and increased the already existing productivity gap between the
manufacturing industry and the service sector – a phenomenon that was later called “Baumol’s disease”.
Not all services sectors, however, are affected similarly by ICT services sometimes noted as an example of
a sector that may contribute to improved productivity growth, i.e. see Wölfl (2003, 2005) for an overview.
For data covering the 1980s, Siegel and Griliches (1992) investigated if outsourcing of services by
manufacturing industries led to an overstatement of manufacturing productivity. However, when
examining the acceleration of manufacturing TFP and outsourcing of services they found only a weak link.
While replicating these results, a later study by ten Raa and Wolff (2001) also provided indirect evidence
supporting the theory of Baumol’s disease. Their findings suggest that manufacturing industries had been
especially successful in outsourcing relatively inefficient services over 1987-1996, but that increasing
services inputs observed alongside with the TFP growth recovery over 1977-1987 could stem from both
outsourcing and a general substitution of service activities for material inputs.
Fixler and Siegel (1999) provide some insights into outsourcing and its productivity impact on the
services sector. Their empirical evidence suggests that outsourcing led to short-run reductions in services
sector productivity, but that productivity improvements can be expected, especially for business services,
once outsourcing of services by manufacturing firms will subside relative to production capacity in the
services sector. They also argue that productivity in the services sector will increase as outsourcing by
service firms increases, although they provide no direct evidence of this.
Over the second half of the 1990s, the United States experienced a strong economic expansion and an
increase in productivity growth. There is little dispute that this was primarily driven by large IT
investments facilitated by falling hardware prices, and the transformation of economic activities that these
investments brought. Mann (2003) argues that although technical change is the most important driver of IT
hardware price declines, international trade and the globalisation of IT hardware production also played a
major role. More specifically, she estimates that over 1995 to 2002, the international fragmentation of IT
hardware manufacturing led to a price decrease between 10% and 30% of IT hardware. This translated into
a higher productivity growth of 0.3 percentage points per year corresponding to an accumulated
USD 230 billion in additional GDP. Moreover, IT capital deepening and IT diffusion are now fuelling an
increasing demand for IT services and software, which has allowed new business areas and economic
activities to emerge. Mann argues that as offshoring in IT services and software production increases,
prices will fall considerably in these areas too, which in turn will foster further productivity increases as
the economy adopts a higher degree of pervasive computing. Moreover, since the price elasticity of IT
services’ demand is higher than that of IT hardware demand, the potential increase in productivity could
even be greater than it was in the 1990s.
That offshoring leads to such beneficial outcomes is also the view of the McKinsey Global Institute
(MGI) (2003) which looked at savings incurred by offshoring customer services to India, and the overall
economic benefits enjoyed by the United States from such investments. Acknowledging that direct labour
cost savings can exaggerate the potential benefits of offshoring, and including related costs such as
increased communication and travel expenses in their estimates, they predict a reduced cost base of
between 45% and 55%. If the process design of the company is re-engineered at the same time, they
envision additional savings bringing the cost base down to 30-35%.8
MGI also estimates a base saving of USD 0.58 per corporate dollar invested in offshoring, and a
directly related benefit to the US economy of USD 0.09 per dollar due to additional exports to India and
profits transfers by India-based US providers. Estimating additional benefits of USD 0.45 to 0.47
stemming from re-employment of workers who lost their job in the process, they estimate an overall
economic benefit to the United States of between USD 1.12 and USD 1.14 per corporate dollar spent
8. The methodology of these estimates is not explained by the McKinsey Global Institute which leaves the
measures open to criticism.
(updated with 2005 figures the benefit is estimated between USD 1.14 and USD 1.17). Benefits to the
Indian economy are estimated at USD 0.33 per dollar, leaving the global economic benefit between
USD 1.45 and USD 1.47.9
According to more recent research by MGI (2004, 2005) the economic benefits of offshoring in the
United States are not the same as those in Europe. In Germany (MGI, 2004), for instance, it is estimated
that the economic benefits are only USD 0.80 per corporate dollar spent on offshoring. This result is
largely driven by lower company savings stemming from the fact that the major German offshore location
is Eastern Europe, and not India where wages are much lower. The economic benefit from re-employment
is also considered lower due to lower flexibility of the labour market in Germany. France is estimated to be
in a similar situation as Germany but with a slightly higher return of USD 0.86 to the economy per
corporate dollar spent on offshoring (MGI, 2005).
Investigations of offshoring such as those by Mann (2003) and McKinsey Global Institute (2003,
2004 and 2005) are useful for getting a clearer picture of the issue, but can lack credibility due to their
underlying estimation methodologies. The US return to corporate offshore investments of about 13% as
estimated by MGI, for example, depends crucially on a significant benefit resulting from re-employment of
labour. While the argument that re-employment is important for reaping economic benefit of offshoring is
noteworthy, the actual estimates could be questionable – especially in the light of a recent BLS study
which found that in the period 1979-1999 31% of those who lost a job due to trade were not fully
re-employed. Besides, while only 36% of the displaced workers were able to find a new job with matching
or higher wages, 55% were at best working for 85% of their former wages, and 25% were working for 70%
or less of their former wage. Also, a recent study by the OECD (2005b) finds that labour adjustment costs
in both the United States and Europe appear to be higher in the industries that are faced with the most
intense international competition. Here, long-term unemployment and losses due to early labour market
retirement following displacement are important issues in Europe. In the United States, on the other hand, a
significant part of the losses accrue to wage losses following the post-displacement job – a finding that
supports the BLS findings.
Other concerns regarding the MGI estimates are raised in Bivens (2005). First, the 13% US return rate
to offshoring is based on proprietary data on firms that have already engaged in offshoring activities. This
leaves room for a potential bias due to self-selection. Second, MGI does not account for the financing
element of the increased imports that offshoring consequently will cause. And last, but not least,
widespread offshoring activities could increase foreign productivity in sectors where offshoring countries
are net exporters, which could result in a loss of income through negative impacts on the terms of trade
(i.e. higher prices for imports relative to prices for exports).10 This argument has also been advanced by
Samuelson (2004), but is downplayed by Bhagwati et al. (2004) by pointing to historical comparisons.
Similar concerns were raised by Europe in the 1950s about growth in the United States, and in the 1970s
by the United States about growth in Japan.
Concerning Mann’s study Bivens notes that although offshoring of IT services has increased, software
prices have fallen more slowly since 2000 than during the previous 20 years, which contradicts Mann’s
argument. Moreover, the immediate effect of increased software investments might not be as significant to
the economy as hardware investments due to a slower depreciation rate of software relative to hardware.
Basing the argument that offshoring of IT services can contribute to productivity growth in the
9. See footnote 8.
10. Other arguments concern the fact that MGI estimates cost savings to be higher than re-employment
benefits. This essentially implies a static redistribution of income from workers to companies per corporate
dollar spent on offshoring. Workers may recoup some of their loss as consumers through dynamic effects
such as lower prices, but this aspect of offshoring remains largely unknown.
US economy on data from the hardware industry could be questioned. Even if this were the case, there
could be further problems. For instance, following Mann’s own estimates and assumptions, Bivens
attempts to re-estimate the productivity impact stemming from a 20% reduction in IT hardware prices. He
finds than in order to arrive at Mann’s estimate of 0.3 percentage points in the yearly contribution to
productivity growth from this source requires a capital share of IT hardware in the US economy about five
times greater than the generally accepted estimates.
One of the first papers to engage in a more detailed analysis of offshore outsourcing and its impact on
productivity was by Egger and Egger (2001b, published in Economic Inquiry 2005). They studied the
impact of outsourcing on the productivity levels of low-skilled workers using data on 22 manufacturing
industries (2-digit NACE) in 12 EU countries over the period 1992-1997. Using the narrow definition of
foreign outsourcing and estimating a production function with CES properties, they found that in the short
run11, a one percentage increase in the outsourcing intensity would lead to a 0.18% decrease in labour
productivity of low-skilled workers. In the long run, which was estimated by excluding specific effects
from the regression, this effect was reversed to a 0.53% increase.
This effect is to some extent supported by Siegel and Griliches (1991). They use data for 4-digit
US manufacturing industries and find a negative correlation between productivity growth and the change
in the share of imported materials although this turns out to be insignificant. The negative short run effect
could also be biased downward due to data measurement errors for capital services (measured by capital
stocks) and labour inputs (measured by employment headcounts). While the former leads to a downward
bias of the capital coefficient, the latter omits the volume and quality of the hours worked. These are wellknown problems and apply to most empirical research in this area (see Siegel and Griliches, 1991).
Moreover, the estimation framework used by Egger and Egger differs from what is commonly used, and
their results are therefore not directly comparable with most other research.
The difference between negative productivity effects in the short run, and positive effects in the long
run, could possibly be explained by short run rigidities, and could, at least in the context of 12 European
countries, be associated with rigid labour markets. For instance, if production is shifted abroad, then for a
given level of employment, labour productivity must fall.
In contrast to Egger and Egger (2001b), Amiti and Wei (2004b) focus on general labour productivity
and not the productivity of the low-skilled, and analyse the impact of offshoring on growth instead of
levels. Using data for 96 US manufacturing industries (2-digit data from the Bureau of Labor Statistics),
their analysis takes form through a more standard framework as they apply a Cobb-Douglas production
function. Offshore outsourcing of materials is distinguished from that of services, which includes inputs
from telecommunications, insurance, finance, business and computing and information services. The
sample runs over the period 1992 to 200112, and as a separate analysis Amity and Wei also look at how
offshoring affects changes in labour demand. The measure of outsourcing is similar to the broad measure
used by Feenstra and Hanson (1999).
Amity and Wei (2004b) find that there is no clear effect on productivity from material offshore
outsourcing, but that there are large positive effects from offshore outsourcing in services. Depending on
the model specification, an increase of one percentage point in services outsourcing intensity leads to an
increase in labour productivity from 0.43 to 0.57 percentage points.13 It should be noted that in a more
11. This was estimated by including time-, industry-, and country-specific effects in the regression equation.
12. To calculate outsourcing for the year 2001, the input/output coefficients from the year 2000 are used.
13. GMM estimates carried out to reflect concerns about endogeneity gave almost identical results as OLS, and
inclusion of lagged dependent variables plus industry fixed effects did not change the signs.
recent study which addresses the potential endogeneity issues raised earlier, Amity and Wei (2006) find
that material offshoring in fact does have a significantly positive, yet small, productivity impact.
A significant contribution from the paper by Amiti and Wei (2004b) is that they investigate effects of
outsourcing over both the aggregated sample of 96 manufacturing industries as described above, but also at
a more disaggregated level (450 industries). This proves to be an important distinction for changes in
labour demand, although this approach can unfortunately not be applied to labour productivity due to lack
Most studies of productivity impacts of offshoring at the aggregate level have focused on large
countries or economic regions such as the United States or the European Union. Compared with smaller
countries, however, these tend to be less open in terms of trade flows, and the effects of offshoring may
therefore also be less revealing – in particular if the data period covered by the analysis is short.
Recognising this, Egger et al. (2001) analyse the productivity impacts of offshoring in Austria using data
for 18 manufacturing industries (2-digit NACE) over 1990-1998. They use a measure of offshoring similar
to the narrow measure employed by Feenstra and Hanson (1999) but are unable to directly distinguish
industry sourcing by the country of origin. As sourcing to the Eastern European countries is particularly
important in the Austrian context, they assume that shares of intermediate imports by country of origin are
equal for all industries.
To establish the impact of offshoring on productivity Egger et al. (2001) follow the approach applied
by Feenstra and Hanson (1999). That is, they examine the change in TFP (measured as a Tornqvist index)14
due to offshoring to Eastern European countries and control for interaction effects with offshoring in
low-skilled as well as capital intensive industries.15 Controls include productivity effects stemming from
R&D expenditures, general import and export openness with Eastern countries measured as the import and
export ratios of gross production, as well as time-specific and industry fixed effects. Egger et al. find that
offshoring has a positive and significant effect on TFP and that this effect seems to be more pronounced
with respect to offshoring in capital intensive industries relative to low-skilled intensive industries. They
estimate that 0.2 percentage points of the 0.9% average increase in Austrian TFP can be attributed to
offshoring. They also found that offshoring to other OECD countries had a significant negative effect on
TFP, although the authors do not discuss this effect.
5.2. Productivity effects at the level of the firm
Aggregate data contribute to the understanding of international outsourcing and its impact on
productivity. However, given the large heterogeneity in firm behaviour with respect to outsourcing,
investigations at a more detailed level can provide further insights.
5.2.1. Outsourcing and its impact on productivity
Some of the earliest attempts to estimate the effects of fragmentation on plant productivity using
micro-data include Görzig and Stephan (2002), and Girma and Görg (2002, 2004). However, while both
studies focus directly on outsourcing as an explanatory factor of productivity, neither of them make the
distinction between domestic and international outsourcing.
14. The Tornqvist Index is a weighted geometric average of the quantity relatives using arithmetic averages of
the value shares in the two periods as weights. It has been shown than when the phenomenon being
analysed (i.e. TFP) can be represented as a homogeneous translog production function, then the change in
TFP can be represented by a Tornqvist index.
15. Dummies are coded as 1 if low-skilled and capital intensities are higher than the sample averages.
Görzig and Stephan (2002) examine a panel data set of about 43 000 German manufacturing
companies over the period 1992-2000. They estimate firm performance measured by both the returns per
employee, which could be interpreted as productivity, and the return on sales.16 Three different measures of
outsourcing are employed, and are all related to internal labour cost. The first is “material inputs” which
reflects make-or-buy decisions by the firms and therefore resembles material outsourcing to external
suppliers. The second is “external contract work” meant to reflect subcontracting, and the third is “other
costs not related to production” capturing outsourcing of services.17 The authors estimate both a
between-firm specification, where all observations are averaged for each firm, and a within-firm
specification, where they control for unobserved heterogeneity and exclude all time-invariant variables.
The former model is interpreted as the long-run model, whereas the latter is interpreted as the short-run
Generally, Görzig and Stephan find a positive and significant effect of all three measures of
outsourcing on firm performance, measured as returns per employee. This effect is strongest for material
outsourcing, but negative for services in the short run. Moreover, they find that increased subcontracting
and outsourcing of services reduces firm profitability, whereas firms engaged in material outsourcing tend
to do better than those that do not outsource. Based on this they conclude that, on average, the level of
subcontracting and outsourcing of services that firms have engaged in is above the optimal level. Despite
the significance of these results, it should be noted that most of the variation in performance across firms
can be attributed to firm-specific characteristics, and therefore remains unexplained.
Girma and Görg (2002, 2004) address the impact of outsourcing on both labour productivity and TFP
for three separate UK manufacturing industries over the period 1982-1992. They use plant-level data18,
including larger establishments with more than 100 employees, in the chemical, electronic, mechanical and
instrument engineering industries. Using the standard framework outlined in the methodology section they
analyse how both productivity levels and growth are affected by outsourcing, although they do not
distinguish between the domestic and international direction of the activity. Outsourcing is measured as the
cost of industrial services received by an establishment relative to the total wage bill, where industrial
services include machine maintenance services and engineering and drafting services only.
compared to other studies, Girma and Görg apply a more narrow measure.
The impact of outsourcing on productivity levels is positive and significant for plants in the chemical
and engineering industries, and about three times stronger in the latter. The effect of outsourcing in the
electronics sector is negative but insignificant. Girma and Görg also include an interaction term for
outsourcing and plant ownership and find that the productivity effect is more pronounced in case of foreign
ownership, especially for plants in the engineering sector. There are no clear results regarding
outsourcing’s impact on the growth of labour productivity, although there appears to be a correlation in
foreign owned plants in the engineering sector. Regarding the analysis of TFP, the sectoral pattern of
effects on labour productivity is repeated. This implies that there is a positive productivity effect of
outsourcing to plants in the chemicals and engineering sectors, but not in the electronics sector, and the
effect for plants in engineering is stronger than that for chemicals. The only exception is that the effects of
16. Both measures make use of the gross operating surplus (GOS). The first is GOS divided by employment,
the second is GOS divided by gross production.
17. The measure “other costs not related to production” may overestimate actual outsourcing of services since
it includes some costs not related to outsourcing activities.
18. Around 80% of their sample is plant level data. The remaining 20% is data collected at the level of the
19. Non-industrial services are not included.
outsourcing generally are even stronger in engineering as well as positively and significantly correlated
with the growth in TFP. Foreign ownership strengthens the effects of outsourcing on TFP.
5.2.2. Offshore outsourcing and its impact on productivity
Using data for larger electronics firms (employment larger than 20) in the Republic of Ireland over the
period 1990-199521, Görg and Hanley (2003b) estimate the effect of outsourcing on both the level and
growth in labour productivity. The data includes a total of 652 establishments covering 12 sub-sectors of
the electronics industry and includes services such as software development, software production,
telecommunications and IT services. Outsourcing is measured as the ratio of imported inputs over total
inputs, and apart from looking specifically at offshore outsourcing they also distinguish between
outsourcing of materials and services. Services are defined as “other direct and indirect costs” excluding
materials, wages, rent interest payments and depreciation. In their analysis they apply the standard
estimation framework outlined in the methodology. To account for possible productivity convergence
when estimating productivity growth, the set of explanatory variables includes lagged labour productivity.
Görg and Hanley find no clear productivity impact of offshore outsourcing in either materials or
services – unless the sample is split into sub-sectors of plants operating either upstream or downstream. In
this case, services outsourcing has a large positive and significant impact on both the level and growth in
labour productivity for plants operating in downstream sectors. A one percentage point increase in
outsourcing is associated with an increase of labour productivity of 0.99 and 0.55 percentage points for
levels and growth rates, respectively. For plants operating in the upstream sectors of the electronic
industry, on the other hand, the effect is negative, albeit insignificant. There is no effect of foreign material
outsourcing in either of the two sub-sector classifications.
Using the same data as above, Görg and Hanley (2005) extend their study of offshore outsourcing in
the Irish electronics sector and its productivity impacts. In this study, however, they focus on TFP instead
of labour productivity and control for differences in export intensities across plants. To do this they
introduce a dummy that takes the value 1 if the plant’s export intensity is higher than the median. Since
they also code for plant-specific effects, the export intensity dummy captures productivity effects for plants
moving from low-export to high-export intensity. Time effects are also included, but they do not
distinguish between upstream and downstream operations as in Görg and Hanley (2003b). The dependent
variable is plant-specific TFP levels.
Looking at offshore outsourcing of materials and services combined, Görg and Hanley (2005) find a
positive and significant effect on TFP. When distinguishing between offshoring of services and materials,
however, materials are found to have a positive and significant productivity impact while the coefficient on
services is positive but insignificant. These effects are robust across different model specifications. There
is no significant productivity impact from the export intensity dummy.
20. Based on their estimates, Girma and Görg also calculate “back-of-the-envelope” contributions of the
change in both labour productivity and TFP resulting from actual changes in outsourcing intensity as
implied by their model. These are then related to the actual changes in productivity. For domestic plants in
the chemical sector, they find that outsourcing contributed 4.7% to the growth in labour productivity and
foreign plants 2.4%. In the engineering sector, the corresponding contributions were larger with 14.7% and
6.8%, respectively. For growth in TFP in chemical plants the contribution from domestic plants was 1.1%
and 0% for foreign plants, whereas the contribution to TFP in engineering was 24.4% for domestic plants
and 0% for foreign plants.
21. The manufacturing of electronics in the Irish economy was a rapidly expanding sector over this period, and
has witnessed a significant fragmentation of production over the past decades (Ruane and Görg, 2001).
Prior to the two studies mentioned above, Görg and Hanley (2003a) conducted another but closely
related study. This was based on the same data source, but addressed the question whether or not
outsourcing increased plant profitability. In the study they find evidence of a causality going from
outsourcing to profitability; they also find that outsourcing of services lowers plant profitability. The effect
on profitability from outsourcing of materials was positive but insignificant. Reversing the regression, they
found that material outsourcing was positively and significantly related to profitability, whereas services
outsourcing was negatively related, but insignificant. It should be mentioned, however, that no plant
specific effects were estimated, which potentially could bias the results. Moreover, they did not distinguish
between national and international outsourcing.
Görg, Hanley and Strobl (2004) extended on Görg and Hanley (2003a, 2003b) by investigating the
effects of offshore outsourcing on plant productivity over a longer time period (1990-1998), and by using
data covering all manufacturing industries. Again, outsourcing of services is distinguished from material
outsourcing, and productivity effects are estimated for productivity levels only following the standard
framework. The only notable difference from Görg and Hanley (2003b) besides larger data coverage is a
slight change in the outsourcing variable which here is measured as the ratio of imported inputs over the
plant’s total wage bill, and not total inputs. The measures of material and services inputs did not change.
Just as in Görg and Hanley (2003b), Görg, Hanley and Strobl (2004) generally find that there is no
productivity effect from services outsourcing. Despite the similarity in estimation frameworks between the
two studies, however, the effect from material outsourcing is both positive and significant, and they
estimate that an increase in material outsourcing by one percentage point leads to a 1.2% increase in labour
productivity. When allowing different production functions according to plant ownership (foreign or
national) and trade activity (domestic or international, based on the exporting status of the plant), offshore
outsourcing of materials is found to have an impact on plant productivity of a similar magnitude for both
foreign and nationally owned exporting plants. The effect for non-exporting plants, regardless of
ownership, was insignificant. Thus, only plants oriented toward the international market seem to benefit
from offshore outsourcing of materials. Taking into account the international character and ownership of
plants did not result in any impact on productivity from offshore outsourcing in services.
Focusing entirely on offshoring of services, Criscuolo and Leaver (2005) study the phenomenon’s
effect on total factor productivity in about 37 000 UK establishments over 2000-2003. The data is
compiled using three different sources and includes explicit values of the plants’ imported and exported
services.22 While the measure of offshored services does not distinguish between different service activities
one of the data sources contains partial information on services expenditures coding for 39 different
activities provided the value of the transaction is GBP 10 000 or above.
Criscuolo and Leaver (2005) estimate a production function using a somewhat more generalised
version of the standard approach. The offshoring variable, which corresponds to the value of services
offshored relative to the total services purchased by the plant, still takes effect through the technology
factor, but the production function is specified in terms of deviations from the industry mean (measured on
a 4-digit level). Controls include different firm characteristics including indicators on whether or not the
firm is part of a multinational enterprise coding for ownership, e.g. the United Kingdom, the United States,
and “other”. Firms that export services are also accounted for, and the regression includes specific effects
for sectors (4 digit level), regional location and time. Using fixed effects GMM estimation the study finds
that offshoring of services generally has a positive and significant effect on plant productivity. However,
when separating the sample into manufacturing and services firms, offshoring only has a positive
significant productivity effect in the latter group. This result supports most other research done with
22. Imports and exports of “all transactions with individuals, enterprises and other organisations domiciled in
a country rather than the UK”.
respect to manufacturing plants. For the plants in the services sector, a 10% increase in offshoring is
associated with a 0.68% increase in TFP.
Criscuolo and Leaver also look at the productivity impacts of offshoring considering their ownership
form and global engagement. Here, combining data on manufacturing and services plants, they find
indications that offshoring is associated with positive and significant productivity effects for domestically
owned firms (but not for those foreign owned), for firms that are not a part of a multinational enterprise
(but not for those that are), and for firms that export (but not for exporters). These results could suggest
diminishing productivity returns to offshoring, as plants that are already globally engaged may have little
more to gain.
Calabrese and Erbetta (2004) analysed the effects of outsourcing on firm performance in the Italian
automotive suppliers sector over the period 1998-2001. Their sample includes 465 plants from the
Piedmont region in Italy. Contrary to most other empirical studies using micro-data, their methodology is
based on “static” ANOVA analysis of firm performance according to firm categories depending on the
degree of outsourcing. They also apply a “dynamic” ANOVA analysis looking at changes in the
outsourcing categories between 1998 and 2001. Three measures of outsourcing all expressed relative to
total operation costs were used. The first relates to material outsourcing, the second to services
outsourcing, and the third is an integration variable measuring costs of personnel, depreciation and
Labour productivity in the Italian automotive industry has generally been in decline over the period,
yet Calabrese and Erbetta find that material outsourcing seems to have lessened the decline significantly.
This indicates a positive effect on labour productivity from material outsourcing. The effects from
outsourcing of services, however, are less clear. Unfortunately, there is also no distinction between
domestic and international outsourcing, and generally the evidence presented in the paper remains
Mazzola and Bruni (2000) analysed the success of 160 small Italian firms over the period 1979-1992.
Based on firm performance with respect to sales, employment and productivity, they constructed a binary
variable representing an observable indicator of a latent success variable. Using a probit model this
variable was then related to on a range of different indicators related to organisational structure, financial
conditions, market structure as well as sectoral variables. Following this approach they found that firms
that are subcontracting work processes have significantly higher probabilities of success relative to other
firms. As the research objective of the study is not primarily concerned with outsourcing, it should be
noted that the subcontracting variable was only entered as a simple dummy.
In an effort to analyse determinants of different offshoring types, Lui and Tung (2004) also consider
the effects from offshoring on firm productivity. The authors consider outward FDI and export outsourcing
as opposed to import outsourcing which is the focus of most other studies. In this setting, export
outsourcing refers to the non-affiliate offshore production of company exports. The data covers 1 336
exporting manufacturing companies in Chinese Taipei, and includes different firm characteristics obtained
from two different sources dating from 2000 and 2001. Productivity relates to labour and is measured as
the sales per employee relative to the weighted industry average.
Labour productivity is linked in regression analysis to both export outsourcing and FDI dummies
including controls for a variety of different company characteristics. Company size, R&D expenditures,
and export related activities are all accounted for as is the direction of FDI, whether foreign affiliates
operate in upstream or downstream markets, and their age. Liu and Tung find that FDI generally is
associated with a negative productivity impact, although it is positive when directed towards the
United States. Effects from FDI towards China are negative and significant. Export outsourcing, on the
other hand, is found to have a positive and significant impact on both levels and growth in labour
productivity. Linking the decision to engage in export outsourcing to lagged labour productivity, including
various controls, shows a positive and significant effect. It thus seems that more productive firms are
engaged in export outsourcing to a greater extent than less productive firms, and that by doing so they
increase their productivity even further.
5.2.3. Foreign direct investment
Foreign direct investment may be associated directly with offshoring in the form of international
insourcing. For this to be the case, however, the FDI must be accompanied by domestic employment
reductions following the halt of production processes carried out by the foreign affiliate. This can be hard
to identify, and the literature that deals with FDI and its impact on firm productivity and profitability
seldom accounts for these. The results may therefore not reflect productivity effects that stem only from
5.3. Indirect evidence
5.3.1. Determinants of outsourcing
Another perspective on outsourcing and its productivity effects may be obtained by looking directly at
the determinants of outsourcing.
Girma and Görg (2004), who estimated the productivity effect of outsourcing in the chemical,
engineering, and electronic manufacturing industries in the United Kingdom (as discussed above), also
estimated the determinants of outsourcing. Focusing on plant characteristics, they found that high wages
were positively related to outsourcing. As they argue, this could suggest that cost-savings are important in
the firm’s decision to outsource. It could also indicate, however, a specialisation process by skill intensive
plants in which they are outsourcing their relatively low-skill intensive processes.
Kimura (2002) focused specifically on the determinants of subcontracting and used survey data
covering 3 723 large Japanese manufacturers of machinery for the year 1994 and their foreign affiliates (all
firms have more than 50 employees). Although the definition of subcontracting used in the survey is
unclear, it appears to include both materials and services, with no distinction between the two, and involves
some form of long-term contractual relationship between the partnering firms. Also, there is no distinction
between domestic and foreign subcontracting. By using a logit specification to estimate the probability of
using sub-contracting, Kimura finds no significant effect from operating surplus over sales, and firm
performance therefore does not seem to be related to subcontracting.
This finding is in contrast with Tomiura (2004) who investigates determinants of outsourcing using a
slightly more recent and much larger sample. The data is from 1998 and covers 118 300 Japanese firms of
all sizes whose main activity is manufacturing.23 Tomiura is able to distinguish between domestic and
foreign directed activities, but looks only at material outsourcing. As with Kimura’s sample, there is no
distinction between unaffiliated contracts and contracts with own subsidiaries, and he therefore measures
offshoring which includes both international in- and outsourcing. Tomiura finds that more productive firms
tend to be more active in offshoring, and similarly that firms with more labour-intensive production tend to
offshore more. Other factors positively related to offshoring are computer usage intensity, highly skilled
employees, and R&D expenditure per employee. This finding is consistent with labour cost saving
strategies and an increased focus on core competencies. Alternatively, it could suggest that higher
23. For instance, if a firm’s main activity changes from manufacturing to wholesale, as has been the case for
many manufacturers of clothing due to outsourcing, it is no longer in the sample.
technological capabilities are required when engaging in offshoring activities, which would support the
theoretical arguments of Bartel et al. (2005) and, to some extent, Freund and Weinhold (2002).
With the same objective of analysing the linkage between international fragmentation decisions and
productivity, in a later study Tomiura (2005) uses the same data as above but distinguishes firms according
to three international activity categories; i) offshore outsourcing, ii) foreign affiliate ownership (or FDI),
and iii) exporting. A fourth “domestic” category includes all remaining firms not engaged in any of the
first three categories. The vast majority of the firms (more than 90%) belong to this group, whereas only
1% of firms are engaged in all international activities simultaneously.
Tomiura (2005) employs four different measures for productivity; i) labour productivity, ii) an
approximation of TFP24, iii) the firm’s market share in the home country, and iv) the firm size in terms of
output. He then relates productivity to a set of firm-specific dummy variables representing the four
different international activity categories and all combinations thereof. Controls include industry-specific
dummies and a dummy for FDI directed towards Asia. This exercise results in an interesting mix of
productivity effects stemming from different globalisation modes. Firms relying only on exports in their
international engagement tend to be among the least productive in the group of globalising firms.
Moreover, while firms engaged in offshore outsourcing are not more productive than exporting firms in
terms of labour productivity, they are so in terms of TFP. This reflects the fact that outsourcers tend to be
relatively labour intensive companies. As regards FDI, firms that invest in Asia tend to be more productive
than both domestic and exporting firms, but less productive than firms directing their investments to the
rest of the world. Finally, firms engaged in more than one, or several, globalisation modes simultaneously
tend to be among the most productive companies regardless of the productivity measure.
5.3.2. Offshoring and labour market issues
Valuable insights in the link between offshore outsourcing and productivity may also be drawn from
research studying the phenomenon’s impact on labour markets. In this regard, the effect on skill-intensity
is of particular interest since a higher skill-intensity may be associated with higher labour productivity,
under some reasonable assumptions.25 In this connection, however, one must be careful with interpretation
as the skill-intensity is most commonly proxied by the non-production workers’ share of the total wage
Evidence from aggregated data
Feenstra and Hanson (1999), who base their research on Leamer (1998), provide an important study
of offshore outsourcing and its impact on the labour market. The methodology they apply has been the
foundation for many later studies. They investigate the impact of offshore outsourcing on the skill-intensity
of US manufacturing industries over 1979-1990 and find that international outsourcing has a positive and
significant effect on the increase in the non-production workers’ wage share.
Hijzen, Görg and Hine (2003) provide evidence that offshore outsourcing accounts for about half the
increase in the skilled cost share relative to the total wage bill in UK manufacturing industries. The
estimates are based on a sample covering 53 industries over 1982-1998, and the measure of outsourcing is
analogous to that used by Feenstra and Hanson (1999). There is no distinction between services and
24. This approximation is defined by adjusting labour productivity for different capital-labour ratios.
25. Assuming that the marginal product of high-skilled labour is larger than that of low-skilled labour, a
positive effect from offshore outsourcing on a firm’s skill-intensity could indicate a positive effect on
labour productivity of the firm.
material outsourcing, but the skill level is estimated by using SOC codes instead of relying on
Focusing on offshore outsourcing measured as outward processing, Egger and Egger (2001) look at
the change in skill ratios in manufacturing industries over 1995-1997 in the European Union. They find
that the growth rate in high-skilled labour relative to low-skilled labour was about 1.8%, and that
outsourcing could explain about 4% of this change. For import-competing industries, however, outsourcing
could account for about 18% of the change. The impact of outward processing must be considered a
somewhat moderate or conservative measure of offshore outsourcing. This is because the measure only
includes international trade for reasons of processing, meaning goods temporarily exported from the EU to
be processed abroad and re-imported into the EU. Hence, not all intermediate imports are included.
The impact of international fragmentation on ratios of skilled to unskilled labour has been
investigated for the German and Italian economies by Helg and Tajoli (2004). They use a narrow measure
of offshore outsourcing similar to Egger and Egger (2001) above, namely industry trade flow from outward
processing to non-EU countries (relative to industry output). Data on outward processing was derived from
Eurostat, and employment data from the respective countries’ national statistics on 20 manufacturing
industries in Germany and 13 in Italy (2-digit ISIC, rev. 3). Changes in skill-composition are coded using
broad occupation data (managers and white-collar employees vs. labourers and apprentices).
Helg and Tajoli (2004) link the ratio of managers and white-collar workers to labourers and
apprentices on outward processing. They control for industry output, capital intensity, an R&D index and
the skilled/unskilled wage ratio, and include industry-specific and time effects. By doing so, they find a
consistent positive and significant impact of outward processing on skill-composition in Italy. In Germany,
the effect is negative but insignificant. There are indications that capital and skills are substitutes in Italy,
while complementary in Germany. Also, as Germany has been involved in outward processing for a longer
period than Italy, the data period may simply be too short to reveal a skill-composition effect in Germany.
This could also indicate diminishing returns to offshore outsourcing in the manufacturing sector.
Strauss-Kahn (2002) analyses the employment inequality in France using disaggregated industry data
for the 1977-1996 period (60 manufacturing sectors). Offshore outsourcing is coded as a narrow measure
defined as the share of imported industry inputs embodied in production, from the same industry.
Strauss-Kahn finds that offshore outsourcing has a significant negative effect on the share of unskilled
employment, and that such activities account for 11% to 15% of the within-industry shifts towards skilled
employment over 1977-1985, and for 25% over 1985-1993. The main effect however, stems from
skill-biased technological progress, which is consistent with the findings of Feenstra and Hanson (1999).
Evidence from disaggregated data
Head and Ries (2002) use data for approximately 1 000 Japanese multinational enterprises in the
manufacturing sector observed in the period 1965-1990. They estimate the impact of offshore employment
on the non-production workers’ share of the total wage bill using a skill proxy, but their study is directed
toward subcontracting to foreign affiliates and therefore relates to international insourcing and not to
offshore outsourcing. The paper is nevertheless included in this survey as it might give additional insights
into the subject which other studies on outsourcing have not covered. Moreover, the definition of foreign
affiliates is relatively wide as it includes offshoring to establishments with 10% Japanese ownership. The
study estimated the effect of offshoring with both industry level data as well as data on the firm level. The
measure of offshoring changes with the level of analysis but is linked to the employment of foreign
affiliates’ share in total firm employment.
Using a translog cost function Head and Ries find no effects of offshore employment on skill-intensity
when using industry aggregates. At the firm level, however, the effect of offshoring is both positive and
significant. Moreover, they find evidence that the sign of the effect on skill-intensity may depend on the
income level of the country to which the offshoring activity is directed. As such, there is a tendency for
skill-upgrading if the offshoring activity is directed toward relatively low-income countries, whereas this
effect is reversed for offshoring directed toward sufficiently high-income countries.
Geishecker and Görg (2005) use employment data from German manufacturing industries over
1991-2000 to estimate a wage equation. In addition to the standard controls such as age, tenure, marital
status, etc. they include a measure of outsourcing which relates to the broad measure of outsourcing in
Feenstra and Hanson (1999). The measure is thus calculated at the industry level (2-digit NACE) and not at
the plant where the person is employed. The authors distinguish between high- and low-skill intensive
industries, and between high- and low-skilled individuals. They find that whereas outsourcing has a
positive and significant effect on wages for high-skilled individuals, this is only true if the person is
employed in industries with high skill intensities. The effect is reversed for low-skilled individuals where
outsourcing has a negative significant effect on wages, but only for those employed in low-skill intensive
6.1. Direct evidence
There appears to be no distinct pattern regarding outsourcing and its effect on productivity from the
literature presented above (some of the main results are presented in Table 1). The most apparent
conclusion one seems to be able to draw at this stage is that the productivity enhancing effect of the
phenomenon depends critically on the context in which it is happening. It is however possible to outline a
number of indications from the literature.
Table 1. Empirical evidence on outsourcing and productivity
Source Country Industry Period Type of
Aggregate level M S SC
Short-run effect – n/a n/a
Long-run effect + n/a n/a
Labour growth General + ++ n/a
TFP growth General + ++ n/a
Amity & Wei (2004b) United States Manufacturing 1992-2001 Offshore Labour growth General 0 + n/a
Egger et al. (2001) Austria Manufacturing 1990-1998 Offshore TFP growth General + n/a n/a
Plant level M S SC
Short-run effect ++ – +
Long-run effect ++ + +
Labour growth 0 0 n/a
Labour level 0 0 n/a
Upstream firms 0 0 n/a
Downstream firms 0 + n/a
M S MS
n/a n/a +
+ 0 n/a
General + 0 n/a
Exporting firms + 0 n/a
Domestic firms 0 0 n/a
General n/a + n/a
Manufacturing n/a 0 n/a
Services n/a + n/a
Domestic n/a + n/a
Foreign n/a 0 n/a
MNEs n/a 0 n/a
Non MNEs n/a + n/a
Exporter n/a 0 n/a
Non exporter n/a + n/a
Chemicals + / +
Engineering ++ / ++
Electronics 0 / 0
Chemicals 0 / 0
Engineering 0 / +
Electronics 0 / 0
(see Type of out.)
FDI General –
Offshore Export outsourcing +
FDI General –
Offshore Export outsourcing +
R. of Ireland Manufacturing
& services Offshore TFP level
Labour / TFP growth
Labour / TFP level
Görzig & Stephan (2002)
Görg et al. (2004) R. of Ireland Manufacturing
Girma & Görg (2004)
Criscuolo & Leaver (2005) United Kingdom Manufacturing
Görg & Hanley (2005)
Labour level & growth
Lui & Tung (2004) Chinese Taipei
Egger & Egger (2001b)
& services Görg & Hanley (2003b) R. of Ireland
Germany Manufacturing Any Return per employee
EU12 Manufacturing 1992-1997 Offshore Low-skill labour level
+ / +
++ / ++
MS * FO
0 / 0
0 / 0
0 / 0
+ / +
Amity & Wei (2006) United States Manufacturing 1992-2000 Offshore
Note: M = material outsourcing, S = services outsourcing, MS = material + services outsourcing, FO = foreign
ownership. A +(-) indicates a positive (negative) significant effect, whereas 0 indicates insignificant effects. Double
signs indicate that effects are larger relative to single signs for the same study. Not all effects depicted here are
necessarily robust over all model variations in the different studies.
Egger and Egger (2001b) and Amiti and Wei (2006, 2004b) study the effect of offshore outsourcing
on an aggregated level looking at the European Union and the United States, respectively. While Amiti and
Wei find no, or only little, evidence of positive productivity effects from offshore outsourcing of materials
in the United States, Egger and Egger find a positive long-run impact in the EU, but a negative effect in the
short run. The difference in effects between the short and long run in the European case could possibly be
explained by labour market rigidities resulting in a sluggish employment adjustment process. For instance,
if an economic activity is shifted abroad, for a given level of employment, ceteris paribus, labour
productivity falls. In the light of the seemingly contradictory results it must be emphasised that the two
studies are not based on the same methodological framework and that the outsourcing measure is broader
for the productivity study on the United States compared to that on the European Union. The effects are
therefore not directly comparable. Work done by Egger et al. (2001) on Austrian manufacturing industries
using a similar approach as Amiti and Wei, however, showed that offshore outsourcing of materials had a
clear positive productivity impact. This could suggest differences in effects related to the size of the
economic region under analysis. Large economic regions for instance tend to be less open in terms of trade
flows, which may play down the role of offshoring.
Plant-level estimates for Ireland (Görg and Hanley, 2003b) and to some extent the United Kingdom
(Girma and Görg, 2004) show little evidence of a productivity impact from material offshore outsourcing
in the electronics manufacturing industries. Besides suggesting the possibility of sectoral differences as
regards productivity effects from outsourcing, the results could also indicate the existence of diminishing
returns to outsourcing. Material outsourcing is generally of a much larger magnitude than services
outsourcing, and gains to such activities could therefore already be close to its optimum level. Offshore
outsourcing of services, on the other hand, is growing from a much smaller level and is also found to affect
productivity positively, at least in the US manufacturing industries (Amity and Wei, 2004b), for
UK services establishments (Criscuolo and Leaver, 2005), and, under certain conditions, for Irish
electronics firms (Görg and Hanley, 2003b). The data coverage in most studies focusing on the economic
impact from material sourcing might therefore not stretch back far enough to capture the phenomenon’s
productivity enhancing effects.
A third reason for the insignificant effect of material outsourcing could be that the outsourcing
measure includes some final goods, in addition to intermediates inputs, and therefore captures a part related
to import competition which could have a negative influence on productivity. There is some indication of
this as Amity and Wei (2004b) find a negative relationship between the import share and productivity.
Görzig and Stephan (2002); Görg and Hanley (2003b) and Criscuolo and Leaver (2005) all provide
evidence of a positive productivity impact of services outsourcing at the plant level. Görzig and Stephan
find that although the outsourcing effect is negative in the short run, it is positive in the long run. This
finding supports the argument of a rigid European (in this case German) labour market as noted above.
Görg and Hanley (2003b) find that the offshore outsourcing of services only has a positive impact in
downstream operating plants of the industry (electronics). This could indicate that plants in the
downstream sectors tend to outsource services that are relatively low-skill intensive allowing them to focus
on high-skill activities. Plants in upstream sectors, on the other hand, seem to be increasing their focus on
low-skill processes by outsourcing relatively skill-intensive tasks. Thus, assuming that upstream activities
are less skill-intensive than downstream activities, these results support the theory of core competences.
Criscuolo and Leaver (2005) find a positive and significant productivity effect from offshoring of services
in UK establishments, but only for firms within the services sector and not for manufacturing companies.
This is contradictory to the results provided by Amity and Wei (2004b) and to some extent Amity and Wei
(2006), but in line with the results provided by Görg and Hanley (2003b) and Görg et al. (2004). As the
two latter studies are based on micro-data while the former is not, the degree of data aggregation when
assessing the productivity effects from offshoring could be an issue.
When comparing the results from Criscuolo and Leaver (2005); Girma and Görg (2002, 2004);
Görg et al. (2004) and Tomiura (2004) some interesting patterns emerges. The first study finds that
offshoring of services in services sector firms generally effects productivity positively but emphasises that
this impact is significant only for domestically owned companies when compared with those that are
foreign owned. The same holds for non-multinationals (when compared against multinationals) and for
firms that primarily operate domestically (when compared against international operating firms). This is,
however, not the case in Girma and Görg who find that foreign ownership re-enforces the effects of
outsourcing (of materials and services combined) on productivity. Similarly, Görg et al. find that the
productivity effects of material offshore outsourcing are of a similar positive magnitude regardless of
whether the plant is foreign or domestically owned (as long as it is an exporting company) but that there is
no productivity impact if the plant only operates domestically. Tomiura also finds that manufacturing firms
are more likely to obtain productivity benefits from material offshoring when already globally engaged.
From this, it seems that being active on the global scene when offshore activities are initiated is important
for enhancing productivity concerning material sourcing, whereas this does not appear to be the case with
respect to services. In fact, the opposite becomes true and could be an indication of much stronger
diminishing returns to offshoring with respect to services compared to materials. This perspective,
however, is somewhat distorted by the results obtained by Lui and Tung (2004) on manufacturing firms in
Chinese Taipei. They find that productivity effects stemming from material offshoring are greatest for
firms that are less globally engaged, but their offshoring measure is narrower as it includes export
outsourcing only and comparisons are therefore not directly applicable.
6.2. Indirect evidence
Following is a summary table of a selection of studies that might help give an indication of the
productivity impact of outsourcing.
Table A. Indirect evidence on outsourcing and productivity
Country Industry Period Type of
Productivity measure Remarks
Feenstra & Hanson (1999) United States Manufacturing 1979-1990 Offshore Non-production cost share growth
Egger & Egger (2001) EU15 Manufacturing 1995-1997 Offshore High-skill share growth
Hijzen (2003) United Kingdom Manufacturing 1993-1998 Offshore Skilled cost share
Hijzen, Görg & Hine (2003) United Kingdom Manufacturing 1982-1998 Offshore Skilled cost share
Germany Manufacturing 1988-1997 Offshore White-collar/blue-collar ratio level
White-collar/blue-collar ratio level
Skilled cost share level
Strauss-Kahn France Manufacturing 1977-1993 Offshore Share of unskilled employment growth
Head & Ries (2002) Japan Manufacturing 1965-1990 Offshore Non-production cost share Multinationals + n/a
High-skilled + 0
Low-skilled 0 –
Germany Manufacturing Individual wages
Geishecker & Görg (2005)
Helg & Tajoli (2004) Italy
Note: M = materials outsourcing, MS = material and services outsourcing, High = skill-intensive industries,
Low = low-skill intensive industries.
For all studies conducted with aggregate data, offshore outsourcing has a positive impact on the skillintensity, proxied by the non-production workers’ or skilled workers’ cost share of the total wage bill, or
the ratio between white- and blue-collar employment. If marginal products are higher for high-skilled
workers relative to low-skilled workers, there is a potential for a positive productivity effect.
There is general support for a potential positive productivity impact in Head and Ries (2002). In
Geishecker and Görg (2005) it appears that offshoring of intermediate inputs and low-skilled labour are
substitutes in low-skill intensive industries, while this is not the case in skill intensive industries. A similar
argument can be made for the complementarity between offshoring and high-skilled labour; there is often a
positive wage effect in skill-intensive industries, but not in low-skill industries. These results could support
the positive productivity effects in skill intensive industries due to increased specialisation and a focus on
core competences, and are thus comparable, in essence, with some of the results derived from Görg and
Hanley (2003b) and, to some extent, Girma and Görg (2004).
In assessing productivity impacts of offshoring it is noteworthy also to draw attention to the
determinants of the phenomenon. Most research in this area finds that firms that are more productive also
are globally engaged to a larger extent (i.e. see Lui and Tung (2004) and Tomiura (2004, 2005)). Thus,
there is potential for an upward bias with respect to the positive productivity effects stemming from
offshoring due to self-selection. Other concerns relate to the fact that companies that engage in offshoring
also tend to be more capital and ICT-usage intensive than those who do not. This could further distort
estimates but could on the other hand also indicate that a certain capacity is required to benefit from global
engagement. This might also help explain why the majority of micro-level studies on offshore outsourcing
attribute far most of the variation in productivity across firms to unobserved heterogeneity.
6.3. Survey indications
As shown in the sections above, the evidence of positive productivity effects stemming from offshore
outsourcing of materials and services does not appear to be overwhelming. The large benefits of engaging
in such activities may therefore be overrated, and much seems to depend on the heterogeneity between
firms. A number of recent surveys support these views to some extent.
One survey covering 5 231 executives across North America and Europe, who would generally be
considered buyers of offshore outsourcing services, found that while 47% of those engaged in offshoring
activities realised cost savings, only 17% did so significantly (above 20%), while 25% had no savings, and
28% experienced increased costs. The average cost reduction accrued to offshoring was slightly below
10% (Ventoro, 2004). The same survey also found that one in three executives admitted moving work back
from an offshore location due to performance problems. Gartner, a consultant firm, also found that
companies which employ outsourcing firms (either domestic or international) for customer service
processes pay 30% more than what the top quartile global companies pay to do the same functions inhouse.
In light of the debate that exists around offshoring and outsourcing and their potential benefits to
firms these survey results are surprising. Other surveys, however, point in another direction. A narrower
survey of 275 finance executives from a broad range of companies conducted by the CFO magazine found
that offshoring of accounting and financial activities led to significant savings in 42% of the cases, whereas
only 10% had no savings. In the automotive industry, more than half of the respondents engaged in
business process outsourcing (BPO) indicated cost reductions of more than 30% (A. T. Kearney, 2003),
whereas a survey of companies primarily in the financial services sector found 48% enjoying the same rate
of savings, albeit from a sample of only 38 companies (Weissman et al., 2003).
With reservations about the quality of survey estimates, response rates and possible selection biases,
one can at least draw two important messages from the above. First of all, although there is an impression
that offshoring brings large economic benefits, it is not clear cut how and if such benefits are indeed
realised. Second, the large cost reduction differences that exist from survey to survey could indicate large
disparities between the realised benefits depending on the activities that are offshored as well as the
sectoral- and firm-specific characteristics.
In practice, it seems that few companies have the organisational and capital capacity to reap the full
potential of internationalisation, and in this light offshoring remains a big-company phenomenon at its
current stage. There is some support for this view. In a quarterly survey by TEC International, a CEO
consulting company, only 5% of about 1 100 mid-size US company CEOs said they intended to outsource
IT jobs overseas. Only 12% said they planned to offshore manufacturing jobs, and about 73% had no
intention of engaging in offshore outsourcing (TEC, 2004). Another survey conducted by Gartner
including 956 chief information officers (CIO), of which half were from North America, draws a similar
picture. Here nearly 80% did not see IT outsourcing or offshore outsourcing as a priority now or in four
years. The Ventoro (2004) survey mentioned above also found that while only 19% of all companies had a
current offshore strategy, this share increased to 95% if looking only at Fortune 1 000 companies. Thus,
since the bulk of firms in the OECD economies are small and medium-sized companies, and assuming that
these surveys give a representative measure of actual future offshore tendencies, the magnitude of
offshoring may not be as large as is often portrayed.
Given the attention offshore outsourcing currently demands in the public debate, surprisingly little
research on the subject exists. The growing trend in offshoring of services, and the number of jobs
potentially affected in Western economies, has directed the main focus towards labour market concerns. As
such, most studies address offshore outsourcing from a labour market perspective, while the phenomenon’s
impact on productivity has been largely overlooked. It is the objective of this paper to give an overview of
the existing research on the productivity effects of offshore outsourcing and its main results.
Most research addressing the link between offshore outsourcing and productivity is based on
micro-data with a strong focus on manufacturing establishments. More recent research, however, is
directed more toward both distinguishing between offshoring of materials and services as well as with a
growing focus on companies in the services sectors.
There are no clear patterns as to how outsourcing affects productivity, and much seems to depend on
both sector- and firm-specific characteristics. However, it is possible to draw some indicatory conclusions
from the literature. First of all, there appear to be signs of diminishing returns to offshore outsourcing of
materials in manufacturing companies as the productivity effects from such activities generally are found
to be either small or insignificant. Contrary to offshoring of services, companies have long been engaged in
offshoring of materials and the benefit from material offshoring might therefore already have reached its
saturation point. Secondly, offshoring of services generally appears to have no productivity enhancing
effects if undertaken by manufacturing companies while the opposite is true for firms in the services sector.
Thirdly, the positive productivity impact associated with offshore outsourcing of materials is generally
more pronounced if the company is already active on the international scene. There are indications,
however, that the opposite is the case for enhancing productivity through offshoring of services which
could suggest rapid diminishing returns to offshoring of services. And lastly, there are several indications
that part of the productivity enhancing effects from material offshore outsourcing is driven by firm-specific
strategic elements such as increasing the focus on core competences.
Amiti, M. and S. Wei (2006), “Service Offshoring and Productivity: Evidence from the United States”,
NBER Working Paper, No. 11926.
Amiti, M. and S. Wei (2004a), “Fear of Service Outsourcing: Is It Justified?” IMF Working Paper,
Amiti, M. and S. Wei (2004b), “Services Outsourcing, Production and Employment: Evidence from the
US”, IMF Working Paper (forthcoming).
Amiti, M. and S. Wei (2004c), “Demystifying Outsourcing, Finance and Development”, December.
Antràs, P. (2003), “Firms, Contracts, and Trade Structure”, NBER Working Paper, No. 9740.
Antràs, P., L Garicano, and E. Rossi-Hansberg (2005), “Offshoring in a Knowledge Economy”, NBER
Working Paper, No. 11094.
Antràs, P., and E. Helpman (2004), “Global Sourcing”, Journal of Political Economy, Vol. 112, No. 3.
Atkinson, R. D. (2004), “Understanding the Offshoring Challenge”, Policy Report, May. Progressive
Baily and Farrell (2004), “Exploding the Myths about Offshoring”, McKinsey and Company.
Baily and Lawrence (2005), “Don’t Blame Trade for US Job Losses”, The McKinsey Quarterly.
Baker, G. P. and T. N. Hubbard (2003), “Make Versus Buy in Trucking: Asset Ownership”, Job Design,
and Information, American Economic Review, Vol. 93, No. 3.
Bardhan, A. D. and C. Kroll (2003), “The New Wave of Outsourcing”, Fisher Center Research Reports,
University of California, Berkeley.
Bartel, A.; S. Lach, and N. Sicherman (2004), “Outsourcing and Technological Change”, NBER working
paper, No. 11158.
Bartelsman, E. et al. (2003), “Comparative Analysis of Firm Demographics and Survival: Micro-level
Evidence for the OECD Countries”, Working Paper, No. 348, Economics Department, OECD, Paris.
Bernstein A. (2004), “Shaking up Trade Theory”, Business Week, December 6.
Bhagwati, J., A. Panagariya, and T.N. Srinivasan (2004), “The Muddles over Outsourcing”, The Journal of
Bivens, J. L. (2005), “Truth and Consequences of Offshoring”, Briefing Paper 155, Economic Policy
Bolton, P. and M. D .Whinston (1993), “Incomplete Contracts, Vertical Integration, and Supply
Assurance”, Review of Economic Studies, Vol. 60, Iss. 1.
Brainard, L. and R. E. Litan (2004), “Offshoring Service Jobs: Bane or Boon – and What to Do?” Policy
Brief, No. 132, The Brookings Institutions.
Calabrese, G. and F. Erbetta (2004), “Outsourcing and Firm Performance: Evidence from Italian
Automotive Suppliers”, 13th Annual IPSERA Conference.
Coase, R. H. (1937), “The Nature of the Firm”, Econometrica, Vol. 4.
Criscuolo C. and M. Leaver (2005), “Offshore Outsourcing and Productivity”, mimeo.
Drezner, W. Daniel (2004), “The Outsourcing Bogeyman”, Foreign Affairs.
Egger, H. and P. Egger (2001a), “Cross-Border Sourcing and Outward Processing in EU Manufacturing”,
The North American Journal of Economics and Finance.
Egger, H. and P. Egger (2001b), “International Outsourcing and the Productivity of Low-skilled Labour in
the EU”, WIFO Working Paper, No. 152 (Economic Inquiry, Vol. 44, Iss. 1, 2006).
Egger, P.; M. Pfaffermayr, and A. Weber (2003), “Sectoral Adjustment of Employment: The Impact of
Outsourcing and Trade at the Micro Level”, IZA Discussion Paper, No. 921.
Egger, P. et al. (2001), “The International Fragmentation of Austrian Manufacturing: The Effects of
Outsourcing on Productivity and Wages”, North American Journal of Economics and Finance,
Vol. 12, Iss. 3.
European Foundation for the Improvement of Living and Working Conditions (2004), Outsourcing of ICT
and Related Services in the EU, Luxembourg: Office for Official Publications of the European
Farrell, D. and V. Agrawal (2004), “Who Wins in Offshoring?”, Global Outsourcing.
Feenstra, R.C. and G. H. Hanson (1999), “The Impact of Outsourcing and High-Technology Capital on
Wages: Estimates for the United States, 1979-1990”, Quarterly Journal of Economics, Vol. 114,
Feenstra, R. C. and G. H. Hanson (2004), “Ownership and Control in Outsourcing to China: Estimating the
Property-Rights Theory of the Firm”, NBER Working Paper, No. 10198.
Fixler, D. J. and D. Siegel (1999), “Outsourcing and Productivity Growth in Services”, Structural Change
and Economic Dynamics, Vol. 10.
Freund, C. and D. Weinhold (2002), “The Internet and International Trade in Services”, American
Economic Review, Vol. 92, Iss. 2.
Garner, A. C. (2004), “Offshoring in the Service Sector: Economic Impact and Policy Issues”, Economic
Review, Third Quarter, Federal Reserve Bank of Kansas City.
Geishecker, I. and H. Görg (2005), “Do Unskilled Workers Always Lose From Fragmentation?”, The
North American Journal of Economics and Finance, Vol. 16. Iss. 1.
Girma, S. and H. Görg (2004), “Outsourcing, Foreign Ownership, and Productivity: Evidence from UK
Establishment-level Data”, Review of International Economics, Vol. 12, Iss. 15.
Görg, H. and A. Hanley (2003a), “Does Outsourcing Increase Profitability”, Working Paper No. 1,
Nottingham University Business School.
Görg, H. and A. Hanley (2003b), “International Outsourcing and Productivity: Evidence from Plant Level
Data”, Globalization, Productivity and Technology, University of Nottingham.
Görg, H. and A. Hanley (2005), “International Outsourcing and Productivity: Evidence from the Irish
Electronics Industry”, North American Journal of Economics and Finance, Vol. 16, Iss. 2.
Görg, H.; A Hanley, and E. Strobl (2004), “Outsourcing, Foreign Ownership, Exporting and Productivity:
An Empirical Investigation with Plant Level Data”, Research Paper 08, University of Nottingham.
Görzig, B. and A. Stephan (2002), “Outsourcing and Firm-level Performance, German Institute for
Economic Research”, Discussion Paper 309.
Griliches, Z. and J. Hausman (1986), “Errors in Variables in Panel Data, Journal of Econometrics”,
Groshen, E. L. and S. Potter (2003), Has Structural Change Contributed to a Jobless Recovery? Current
Issues in Economics and Finance, Federal Reserve Bank of New York, Vol. 9, Nr. 8.
Grossman, G. and E. Helpman (2002), “Managerial Incentives and the International Organization of
Production”, NBER Working Paper, No. 9403.
Grossman, G. and E. Helpman (2002), “Outsourcing in a Global Economy”, NBER Working Paper,
Head, K. and J. Ries (2002), “Offshore Production and Skill Upgrading by Japanese Manufacturing
Firms”, Journal of International Economics, Vol. 58, Iss. 1.
Helg, R. and L. Tajoli (2005), “Patterns of International Fragmentation of Production and Implications for
the Labour Markets”, North American Journal of Economics and Finance, Vol. 16, Iss. 2.
Heshmati, A. (2003), “Productivity Growth, Efficiency and Outsourcing in Manufacturing and Service
Industries”, Journal of Economic Surveys, Vol. 17, Iss. 1
Hijzen, A. (2003), “Fragmentation, Productivity and Relative Wages in the UK: A General Equilibrium
Approach”, Royal Economic Society Annual Conference 2003, No. 108.
Hijzen A.; H. Görg and R. C Hine (2003), “International Fragmentation and Relative Wages in the UK”,
IZA Discussion Paper No. 717.
Hijzen A. H. Görg and R. C Hine (2004), “International Outsourcing and the Skill Structure of Labor
Demand in the United Kingdom”, IZA Discussion Paper No. 1249.
Huws, U. (2004), EMERGENCE: eWork in a Global Economy, www.emergence.nu
Hummels, D. et al. (2001), “The Nature and Growth of Vertical Specialization in World Trade”, Journal of
International Economics, Vol. 51.
Kimura, F. (2002), “Subcontracting and the Performance of Small and Medium Firms in Japan”, Small
Business Economics, Vol. 18, Iss. 1-3.
King, J. (2003), “IT’s Global Itinerary: Offshore Outsourcing is Inevitable”, ComputerWorld,
Kirkegaard, J. F. (2004), Outsourcing – Stains on the White Collar?, Institute for International Economics.
Kletzer, L. (2001), Job Loss from Imports: Measuring the Costs, IIE, September.
Liu, B. J, and A. C. Tung (2004), “Export Outsourcing and Foreign Direct Investment: Evidence from
Taiwanese Exporting Firms”, Dynamics, Economic Growth, and International Trade, conference
Loecker, J. K. De and J. Konings (2004), Creative Destruction and Productivity Growth in an Emerging
Economy: Evidence from Slovenian Manufacturing, No 4238, CEPR Discussion Papers.
Maliranta, M. (2003), Micro Level Dynamics of Productivity Growth: an Empirical Analysis of the Great
Leap in Finnish Manufacturing Productivity in 1975-2000.
Mann, C. L. (2003), “Globalization of IT Services and White Collar Jobs: The Next Wave of Productivity
Growth”, International Economics Policy Briefs, PB03-11.
Masso, J., R. Eamets, and K. Philips (2004), “Creative Destruction and Transition: The Effects of Firm
Entry and Exit on Productivity Growth in Estonia”, IZA DP No. 1243.
Masten, S. (1984), “The Organization of Production: Evidence from the Aerospace Industry”, Journal of
Law and Economics, Vol. 27.
Mataloni, R. J. (2004), “A Note on Patterns of Production and Employment by U.S. Multinational
Companies”, Survey of Current Business 84(3), U.S. Bureau of Economic Analysis, March 2004:
Mazzola, F. and S. Bruni (2000), Journal of Economic Behavior and Organization, Vol. 43.
McKinsey Global Institute (2003), Offshoring: Is It a Win-Win Game? McKinsey and Company.
McKinsey Global Institute (2004), Can Germany Win from Offshoring? McKinsey and Company.
McKinsey Global Institute (2005), How Offshoring of Services Could Benefit France, McKinsey and
McMillan, J. (2004), “Quantifying Creative Destruction Entrepreneurship and Productivity in New
Zealand”, Motu Working Paper, No. 04-07.
Minevich, M. and F.-J. Richter (2005), Global Outsourcing Report, HORASIS, Geneva.
OECD (2005a), Aspects of Offshoring and their Impact on Employment: Measurement Issues and Policy
Implications, DSTI/IND(2005)7, internal working document, OECD, Paris.
OECD (2005b), “Trade-adjustment Costs in OECD Labour Markets: A Mountain or a Molehill”, OECD
Employment Outlook 2005, Chapter 1, pp. 23-72, OECD, Paris.
Ono, Y. and V. Stango (2005), “Outsourcing, Firm Size, and Product Complexity: Evidence from Credit
Unions”, Economic Perspectives, Issue Q1.
Ruane, F. and H. Görg (2001), “Globalization and Fragmentation: Evidence for the Electronics Industry in
Ireland”, in Arndt, S. and Kierzkowski, H (eds.), Fragmentation: New Production Patterns in the
World Economy, Oxford University Press, pp. 144-163.
Siegel, D. and Z. Griliches (1992), “Purchases Services, Outsourcing, Computers, and Productivity in
Manufacturing”, in Griliches, Z. (ed.), Output Measurement in the Service Sector, University of
Schultze, C. L. (2004), “Offshoring, Import Competition and the Jobless Recovery”, The Brookings
Institution, Policy Brief, No. 136.
Skodzinski, N. (2004), “Offshoring and the Global Marketplace”, Booktech Magazine, September/October.
Slaughter, M. J. (2004), Insourcing Jobs: Making the Global Economy Work for America.
Strauss-Kahn, V. (2002), The Impact of Globalization, through Vertical Specialization, on the Labor
Market: the French Case, New York University.
Ten Raa, T. and E. N. Wolff (2001), “Outsourcing of Services and the Productivity Recovery in US
Manufacturing in the 1980s and 1990s”, Journal of Productivity Analysis, Vol. 16.
Tomiura, E. (2005), “Foreign Outsourcing, Exporting, and FDI: A Productivity Comparison at the Firm
Level”, Discussion Paper Series No. 168, Kobe University.
Tomiura, E. (2004), “Foreign Outsourcing and Firm-level Characteristics: Evidence from Japanese
Manufacturers”, Discussion Paper Series, No. 64, Hitotsubashi University Research Unit for
Statistical Analysis in Social Sciences.
van Welsum, D. and G. Vickery (2004), Potential offshoring of ICT-intensive using occupations,
DSTI/ICCP/IE(2004)19/FINAL, OECD, Paris.
Wölfl, A. (2005), The Service Economy in OECD Countries, STI Working Paper, 2005/03, OECD, Paris.
Wölfl, A. (2003), Productivity Growth in Service Industries: An Assessment of Recent Patterns and the
Role of Measurement, STI Working Paper, 2003/7, OECD, Paris.