Submited paper

by user

Category: Documents





Submited paper
Business Services Offshoring and Value Chains:
Gauging the Consequences
Barbara Dluhosch∗
Thorsten Hens‡,♦
February 14, 2013
Abstract. Recently, advances in information and communication technologies
(ICT) gained economy-wide importance, raising concerns that neither services nor
high-skilled labor might be sheltered from international competition but increasingly susceptible to offshoring. The paper presents a novel theoretical framework for analyzing its consequences with a focus on skilled labor in managing
value-added chains. Accordingly, technological advances in business services and
even North-North trade integration do indeed affect production and specialization patterns. However, although both give rise to a productivity-boosting effect
promoting scale in value chains due to firms sharing in the fixed costs of businessservices production, implications differ in several respects – including offshoring
and policy options.
JEL-Classification: F16, F12, L89, J31
Keywords: International Trade, Value Chains, Service Links, Business Services,
Labor Markets
Department of Economics, Helmut Schmidt University / University FAF Hamburg,
Holstenhofweg 85, D-22043 Hamburg, Germany, email: [email protected] (corresponding author).
Department of Banking and Finance, University of Zurich, Plattenstrasse 32, CH-8032
Zurich, Switzerland, email: [email protected]
Norwegian Business School, Bergen, Norway.
According to a great many indicators, services become increasingly important, such as, for
instance, as a share in value added and, more specifically, as an input in the production
of many (tangible) goods. This also applies to trade. In 2011, commercial services alone
made up for a fifth of world trade and counting (WTO 2012a, p. 22).1 On a value added
basis, that is, acknowledging the fact that many services are incorporated in goods trade,
estimates jump to some 47 percent of all trade (Escaith 2008; Sturgeon et al. 2012 on
measurement issues). Due to the dynamics of technical change, in particular in the ICT
industries (communication, computer and information), all of which are sub-segments of
commercial services trade, prices fell rapidly while at the same time quality went up (OECD
2012). Hence, sheer trade numbers may well underestimate the economic impacts. Advances
in ICT may not only lower the costs of a fragmented and geographically dispersed production
process and thus allow to take advantage of local cost savings in the production of specific
components (see Abramovsky & Griffith 2006 for a quantitative account). Rather, they may
exert a leverage effect on other services which consequently become tradable. This holds in
particular for the category of “other business services” which constitutes the biggest slice in
commercial services trade and mainly comprises inputs to production.2 Higher tradability
might imply that a larger fraction of these business services, as well, is produced offshore.3
That is, services (without government services) on a balance of payments basis. Owing to the 2008 crisis,
numbers in 2011 are still somewhat depressed compared to their pre-crisis levels, even when excluding
finance and insurance.
According to the WTO classification commercial services are composed of transportation, travel and other
commercial services with the latter in turn comprising communication services, construction, insurance,
financial services, computer and information services, royalties and license fees, other business services,
and personal, cultural and recreational services. So-called “other business services” trade added up to
roughly half of all “other commercial services” trade and some 25.7 percent of total trade in commercial
services. Together with computer, information and communication services, “other business services”
made up for 34.4 percent of total trade in commercial services. As far as being related to a geographically
dispersed production process, those services also have an effect on transportation with a share of 20.6
percent in 2011 (WTO 2012b). However, due to their very nature as well as for historical reasons, services
trade is hard to quantify exactly. See Lipsey (2009) for a discussion; for a survey on the role of trade
policy besides technology see Francois & Hoekman (2010).
We follow the convention established since UNCTAD (2004, 148) that defines offshoring as imports of (intermediate) producer services, no matter whether produced at “arm’s length” (international outsourcing)
or in-house but abroad (i.e. captive offshoring or vertical FDI).
Whereas the previous wave of material offshoring primarily affected low-skill intensive components of production,4 this time seems to be different. Business services are mainly produced
by use of skilled labor (Gonzales et al. 2012; Lewin 2011). The tradability thus raises concerns in many developed countries that while previously it was low-skilled labor that was
substituted by cheap imports of intermediates, it might now be skilled labor that is going to
suffer a decline in (local) demand. Popular examples include software engineering or X-ray
examination in India (Friedman 2007). According to guesstimates by Blinder (2006; 2009)
in the US alone 28 to 42 million service jobs might be at risk of being offshored.5
In this paper, we show by means of a North-North trade model that offshorability need not
imply that those services are actually produced offshore nor does offshoring of skill-intensive
services necessarily hurt the high skilled. Rather, effects might be very diverse. The reasons
are threefold. First and foremost, increases in the demand for business services are accompanied by a productivity-boosting effect due to firms sharing in the fixed costs of business
services production. The productivity effect tends to cushion substitution-related pressures
on jobs and wages. While scale and productivity effects of international specialization have
been noted in the literature,6 economic characteristics in business-services production even
Although material outsourcing was almost unanimously considered to affect low-skilled labor in particular
(e.g. Feenstra & Hanson 2003 for a survey), closer examination showed that vertical specialization and
trade had some peculiar effects easily overlooked when transposing the logic of horizontal North-South
trade along HOS lines into the vertical dimension (e.g. Arndt 1997, 1998; Grossmann & Rossi-Hansberg
2008; Dluhosch 2006, 2010; Horgos 2011). Due to these peculiarities, there are well situations conceivable
in which skilled labor might be negatively affected even in case of material offshoring.
See also Blinder & Krueger (forthcoming) for a refined approach offering also more disaggregated numbers
as well as suggestions on appropriate survey design. Based on a case study approach Levy & Murnane
(2005) arrive at similar conclusions as regards the kind of services jobs that are likely to be offshorable;
gauged by the ICT intensity of jobs, Van Welsum & Vickery (2005) estimate that some 20 percent of total
employment is potentially offshorable in the EU15, Australia, Canada and the US. Other, alternative,
approaches to potential offshoring draw on geographic concentration (industry and occupations) within
the US (Jensen & Kletzer 2006; 2010) or domestic US experience as regards outsourcing Kletzer (2009a).
While they arrive at similar numbers with respect to the potential offshorability of jobs, they also stress
that comparative advantage may work in the opposite direction, i.e. towards onshoring, so that effective
numbers in the future might be smaller than those estimates (similar Liu & Trefler 2008). However,
although approaches vary widely, results concerning potential offshorability are remarkably similar in
terms of numbers.
See, e.g., Francois (1990) for an early approach and Grossman & Rossi-Hansberg (2008) more recently.
beef up those effects. Second, most of the trade in business services is North-North. Though
North-North trade is suspected to have less pronounced distributional effects, this need not
carry over to trade in business services. At any rate, effects turn out to be very different from
the usual logic of North-South driven material offshoring. Third, effects depend on the actual
driving forces of this process. Technological advances (e.g. ICT) have effects very distinct
from trade integration. While both boost productivity, technological advances tend to favor
the high skilled, notably, despite (skilled) service offshoring, whereas trade integration tends
to work in the opposite direction.
Since technologies pushing the tradability of services are fairly recent, contributions examining the phenomenon and the implications of (business) services offshoring are still few in
numbers. Though illuminating some of the features of this process, empirical and theoretical
approaches thus far also exhibited quite a number of puzzles which provides the motivation
for rethinking the standard HOS-inspired theoretical framework implicitly underlying most
of those empirical studies. In particular the strong complementarity between offshoring and
skilled employment found by some of the studies is at odds with the conventional substitution perspective on offshoring and the fear of Blinder-type guesstimates. We try to explain
the diverse observations (substitution-effects on the one hand, complementarity on the other
hand) as well as their accompanying symptoms by mapping out a novel framework that allows to trace them back to two different driving forces – thus also hoping to accommodate
the set up of subsequent empirical models trying to disentangle cause and effects.
While guesstimates like those by Blinder were not received without criticism on the account
of aggregate numbers still being comparatively small (Amiti & Wei 2005; 2009; see Crinò
2009 for a survey), evidence of “onshoring” (Jensen 2011; Baldwin 2009) or the requirement
of tacit knowledge in many professional services (Yu & Levy 2010), it can hardly be neglected
that the internet penetrates the provision of a number of (business) services so much that
it fundamentally alters production processes including services in general. Focusing on the
US in the second half of the 1990s, Freund & Weinhold (2002) provide empirical evidence
according to which the effect is not only significant on a statistical account but also economically relevant with a 10-percent increase in internet penetration being associated with more
than a percentage point increase in business, professional, and technical services imports.
Extrapolating these numbers thus suggests that prospects for those employed in commercial
services are bleak indeed.
Retrospective studies, on the other hand, vary widely in the sense that they try to estimate
the actual (as opposed to the potential) effects of (business) services offshoring with respect
to different dimensions of the labor market (aggregate employment, workforce composition
within firms, occupations & wages), different countries and levels of aggregration. Nevertheless, approaches and results share some patterns. Though not exclusively, most studies
on the implications of the actual amount of service offshoring refer to the new competitors
in Asia, in particular India and China (e.g. Kirkegaard 2008; Aspray 2010; Ghani 2010;
Liu & Trefler 2008; 2011). Hence, they approach the issue from a North-South perspective.
Accordingly, overall effects are (still) small, but negatively biased towards the lower skill
groups. Amiti & Wei (2005), however, find no net effect of services offshoring on employment at the aggregate level. At a more disaggregate level of 450 US-industries (4-digit SIC),
though, a statistically significant negative effect comes to the forefront. Accordingly, services
offshoring was responsible for 0.4 of a percent fall in US manufacturing employment in the
years 1992-2000 (Amiti & Wei 2009). Based on US occupational data for the years 1990-2006,
Crinò (2010) presents evidence of services offshoring negatively affecting employment at the
lower end of the skilled-worker distribution. Shifting focus to nine European countries in
1990-2004, Crinò (2012) also seems to lend support to this notion. Studying UK micro-data
from 1992-2004, Geishecker & Görg (forthcoming) find evidence of a significant depressing
effect on real wages of the lower skilled. Prima facie, results thus suggest that effects are
similar to and very much in line with material offshoring along North-South lines – except
for that the effects moved somewhat up the skill ladder. Crinò (2012, p. 20) thus maintains
that “... effects are qualitatively identical, and quantitatively similar, to those of material
Yet, some observations proved to be difficult to reconcile with a pure cost-savings perspective that stresses the substitution between domestic and foreign labor along North-South
lines. One of the puzzling attributes of services offshoring is the unexpected complementarity in occupations and compensation respectively. Figure 1 gives a foretaste.
Figure 1. Business services employment and compensation
Source: OECD, STAN Database, retrieved Feb 01, 2013; own calculations.
The left hand panel indicates an upward trend in US business services employment over
the years 1991-2009. Business services employment trended upwards despite of it becoming
relatively more expensive. Focusing on individual countries and years in 25 OECD countries
for which data is available, the right hand panel shows that in a great many cases covering
the years 1991-2009, the expansion in employment did not come at the expense of wages
(namely in all those situations which are in the upper right hand quadrant). Quite to the
contrary, compensation outpaced average compensation in the respective economies.
Signs of a positive rather than a negative correlation of offshoring and employment prospects
show up in a number of studies. Focusing on services in the years 1977-1999, Harrison et al.
(2007), for instance, find evidence that employment changes of US multinationals firms and
their foreign affiliates in high-wage countries went in tandem. Their findings are in line with
UK firm-level evidence for the years 1997-2005 presented by Hijzen et al. (2011), who focus
on imported services inputs (as a proxy for producer services offshoring) and who find no
evidence of a crowding out of domestic employment thus far but rather a crowding-in effect.
In speculating on how to explain this apparent puzzle, they point out that the vast majority
of this trade takes place with other developed countries rather than developing or emerging
economies which might be key to explaining this symptom.
Hence, while the North-South route certainly explains some of what goes on with respect
to trade in business services, a large(r) part does not quite fit the widely held picture of
it being exclusively low-cost competition from newly emerging economies. Though reliable
numbers on total offshoring of business services in particular are difficult to obtain, the
focus on emerging economies also contrasts with the fact that trade flows in commercial
services seem to be largely a North-North rather than a North-South phenomenon, with the
US, Canada, the EU 27 and Japan making up for 55.8 and 61.9 percent of all imports and
exports respectively. The popular examples of India and China, by contrast, account for 3.3
and 4.4 percent of all exports (7.3 percent including HongKong).7 In addition, numbers have
to be interpreted in light of the fact that in North-South perspective India and China are
usually seen as a unit rather than as very large and diverse countries with myriads of lowpaying unskilled jobs on the one hand and “hot spots” in business-services production with
the latter not only relying heavily on high-skilled labor but also very much geographically
concentrated (Desmet et al. 2012) – to the effect that trade is more of the North-North
A closer look at the empirical studies thus point towards the potential value added a
theoretical analysis might have on the issue, both, for disentangling and thus getting a
WTO 2012b; see also Trefler (2005) for a critique of the predominant North-South perspective. In more
disaggregated perspective, India’s and China’s share in world exports of information, communication &
computer services reached 12.3 and 2.7 percent in 2009 (the latest numbers available), however, with
numbers for India still estimates. Shares of export values in “other business services”, which constitute
a much larger fraction of all trade in commercial services, amounted to 3.2 and 5.2 percent respectively
(WTO database, retrieved Nov. 04, 2012).
According to Hummels, Ishii & Yi (2001), the North-North bias also extends to the narrower measure of
vertical specialization.
better grip on the various channels – and for ultimately helping to sort out cause and effect.
While theoretical work on fragmentation, value added chains, and trade in intermediates has
mushroomed since the 1990s and virtually took off at the turn of the century (e.g. Jones &
Kierzkowski 1990; Deardorff 2001a; Feenstra & Hanson 2003; Yi 2003 to name a few),9 the
vast majority of those studies is geared towards material offshoring. Since producer services
are mostly intermediates, it has been conjectured that the logic of material offshoring readily
extends to the offshoring of business services. Amiti & Wei (2005, p. 317), for instance, weigh
in by stating “[i]n all of these models, the focus is on the outsourcing of material inputs but
these could, in principle, be re-interpreted as service inputs.”
In this sense, models on value chains have been recently refined in order to account for trade
in tasks rather than solely trade in tangible components (e.g. Grossman & Rossi-Hansberg
2008; Baldwin & Robert-Nicoud 2011; Kohler & Wrona 2011). While catering to this notion,
they deviate on two accounts from the empirics mentioned previously. First, they cling to
the North-South perspective in that differences in factor proportions drive many (though not
all) of the results. Grossman & Rossi-Hansberg (2012) are an exception in that they impose
the task-trade framework on two identical countries (in terms of factor proportions) and then
explore how differences in size matter for specialization patterns. However, secondly, all of
them merge business services and the production of other components into tasks with some
of them varying tradability as tasks might be associated with (localized) external economies.
Our approach differs from that in that it clearly distinguishes fragments (that is steps in the
production process itself) and service links managing the value added chain. Clearly separating the two while nevertheless examining their joint effects allows us to capture different
The trend towards slicing up the value added chain also holds within countries. The overwhelming majority
of OECD-countries experienced an increase in the share of intermediates in gross output since 1991, with
the unweighted average of 25 OECD countries climbing by approx. 5 percentage points from 49.96
percent in 1991 to 55.32 percent in 2008 (own calculations based on data supplied by the OECD.stat).
The unweighted average shows that the trend is by no means due to a small number of (large) countries
dominating the picture, but that the development is comprehensive. Exceptions are few: numbers for
New Zealand stayed almost constant, as did those of Portugal; Norway shows a slight decline; only Greece
and the Slovak Republic truly stand out from the crowd with a decline of 1.5 and 3 percentage points
engines in the offshoring of business services and the impacts on domestic occupations in particular. In distinguishing between production of the various parts in the value chain and the
(special characteristics of) business services necessary in managing value added chains our
work is more in line with the intuition of service links along the lines of Jones, Kierzkowski &
Lurong (2005), Deardorff (2005) and related. However, here, we present a novel framework
that allows to track down the impacts of service offshoring on service links in the importing
country.10 While Debaere et al. (forthcoming) also stress that this makes for a difference,
they are not especially geared to business-services trade nor do they explore the substitution versus compementarity issue – and trace them back to different sources in North-North
trade and in the process of services offshoring as we do. Unlike Antràs (forthcoming), Spencer
(2005), and the family of literature that focuses on the peculiarities of principal-agent and
industrial-organization aspects in producing at arm’s length, we abstract from those issues
in order to concentrate on how the different engines in North-North trade matter for one
symptom (co-movement) dominating the other (substitution) and the peculiarities of service
1. Demand in Home and in Foreign
Consider two countries, Home and Foreign, endowed with high and low skilled individuals
H, H and L, L , and with variables of Foreign denoted by an asterisk.11 While allowing for
differences in the size of endowments, we suppose that factor proportions κ ≡ L/H are the
same, i.e. κ = κ∗ . In order to isolate effects rooted in the production of business services
from primarily substitution-driven offshoring via an induced fragmentation of production
and intermediates trade in business services, we thus concentrate on the purest of all NorthNorth channels which is void of any differences in factor-proportions. However, cost-related
On the (empirical) importance to distinguish these two economic activities see Francois & Woerz (2008).
Qua assumption, labor markets are completely segmented by skill level.
substitution will nevertheless play a role, as we will see shortly.
Suppose, moreover, that individuals in both countries share the same preferences over a
homogeneous good and a set of differentiated goods, with one subset of the latter produced
in Home, the other in Foreign. Let the aggregate quantities consumed in Home be denoted
by c0 (homogeneous good) and ci , c˜j (differentiated goods), with index i (i = 1, ..., n) Homemade varieties and index j (j = 1, ..., n∗ ) Foreign-made varieties (c∗0 , c∗i , c˜j ∗ for Foreign).
Quantities as well as varieties, n and n∗ , will be determined endogenously. Hence, we will
track down adjustment at the intensive as well as the extensive product margin in response
to the two drivers of business-services trade. Specifically, we assume the following aggregate
log-linearized utility function U over the three groups of goods
U (c0 , ci , c˜j ) =
µ ln c0 + (1 − µ)  ln ci
ln c˜j 
with µ the expenditure share of the homogeneous good. Qua symmetry-assumption (see
also subsection 2), the quantity ci consumed by Home’s individuals is the same across all
varieties i = 1, ..., n, j = 1, ..., n∗ , i.e. c1 = .. = ci = ... = cn and ci = c˜j ∀i = 1, ..., n;
∀j = 1, ..., n∗ . In the aggregate, demand for differentiated goods by consumers in Home is
thus c =
i=1 ci
= nci and ce =
j=1 c˜j
= n∗ cj (or ci = c/n and c˜j = ce/n∗ for that matter).
Hence, preferences can be rewritten in the form of
c ce
U c0 , , ∗
n n
= µ ln c0 + (1 − µ) (n (ln c − ln n) + n∗ (ln ce − ln n∗ ))
The adding-up properties of the model imply that utility is maximized subject to the aggregate budget constraint
Y = c0 + pc + p∗ ce
with Y aggregate income and p, p∗ the price-index of varieties produced in Home and in
Foreign respectively. Notably, given the preferences in eq.(2), the number of varieties n,
n∗ affect the marginal rate of substitution (MRS): focussing on the trade off between the
homogeneous good on the one hand and the set of locally produced varieties of differentiated
goods on the other hand, eq.(2) implies MRSc0 ,ci = µc/ ((1 − µ) nc0 ) = 1/p. Solving for
p, we obtain p = (1 − µ) nc0 / (µc). Hence, the MRS depends on the number of varieties n
(similarly with respect to Foreign-made varieties) which implies adjustment at both product
margins. First-order conditions entail the following aggregate demand functions in Home
((1 − µ) (n + n ) + µ) p0
(1 − µ) n
ci =
((1 − µ) (n + n ) + µ) np
(1 − µ) n∗
c˜j =
((1 − µ) (n + n ) + µ) n∗ p∗
c0 =
With µb ≡
((1−µ)(n+n∗ )+µ)
c0 = µb
eqs.(4) shorten to
ci =
(1 − µ) Y
and c˜j =
(1 − µ) µb Y
By analogy, we obtain demand in Foreign c∗0 , c∗i , c˜j ∗ , by use of Y ∗ instead of Y , and budget
restriction Y ∗ = c0 p∗0 + pc∗ + p∗ ce∗ .
In principle, national income Y , Y ∗ is composed of labor income and profits π, π ∗ . With
earnings wL , wH of low- resp. high-skilled individuals, labor income in Home is wL L + wH H.
Regards profits, we suppose that the production sector of each economy contains the two
beforementioned types of firms, that is the numéraire and the differentiated goods. While the
numéraire is, as usually, considered to be produced in a constant returns perfect competition
framework with aggregate profits π0 = 0 and with the number of firms indeterminate, each
differentiated good shall be produced by a single firm under increasing returns, so that the
aggregate profits in the differentiated-goods sector are
πi . However, the production of
these goods requires the input of business services for coordinating production. We suppose
that they themselves are supplied by an endogenously determined number m of firms in an
increasing returns framework. If the latter operate under symmetrical cost structures and
face the same demand, they earn profits mπn+1 . Hence, there are actually three types of
firms in the economy, however, with only two of them supplying goods for final consumption.
Business services, by contrast, cater to the production of differentiated goods in that they
help to manage a multistage production process (value chain). Despite the increasing-returns
assumption we will assume that firms in the differentiated-goods and in the business service
sector earn zero profits in equilibrium due to free entry in both businesses. Hence, national
production and income accounts imply
Y = w L L + w H H + π0 +
πi + mπn+1
and Y ∗ = wL∗ L + wH
H + π0∗ +
πj∗ + m∗ πn+1
with π0∗ +
πj∗ + m∗ πn+1
= 0 (Foreign).
2. Supply in Home and in Foreign
The economy has thus three sectors: a numéraire sector, the differentiated goods sector and
a business services sector. Since the supply side of Foreign is thought of as symmetrical to
Home’s except for the absolute endowment with low- and high-skilled labor (however, with
factor proportions identical), we will concentrate on the description of Home’s production
side. Table 1 displays Home’s structure as well as input-output linkages.
< insert Table 1 about here >
In order to account for increasing returns and special economic characteristics in business
services production (with more details below), we assume that the numéraire sector not only
caters output x0 to final demand (c0 ) but also produces inputs X n+1 for setting up production
in the business sector and
total x0 = c0 +
X i for each firm in the differentiated-goods sector, i.e., in
X i +X n+1 , with all of them produced by means of a single factor of
production, low-skilled labor, according to the linear production function x0 = aL0 resulting
in the profit function π0 = aL0 − wL L0 .
By contrast, each line i in the differentiated-goods production requires four sorts of inputs: high- and low-skilled labor, Li and Hi respectively, business services zi and a fixed
amount of the numéraire good, X i , for setting-up production. Business services zi , in turn,
are assumed to enhance the productivity of high-skilled labor Hi thus giving rise to their
ambiguous character: on the one hand, there is a substitutive relationship between the two
of them (domestic demand for skilled labor and business-services offshoring); on the other
hand, there is complementarity between both of them. The complementarity operates via the
productivity-boosting effect as business services help to manage a more fragmented production chain. The latter effect is even beefed up by the sharing of fixed costs in business-services
production and ICT. The two channels will be displayed by two parameters, which will turn
out to be of utmost importance for the net effect of offshoring on skilled labor. The first
of these parameters, that is the fixed costs (X n+1 /m), will be partly endogenous (due to
changes in the number of firms m); the second, that is ICT (δ), will be considered exogenous
(see below).
However, staying for the moment with the production of differentiated goods themselves,
we will presume the following particular function of a representative firm i producing in
Home’s value chain
Ai ≡ Lαi (zi Hi )β
with α, β < 1.12 Business services zi can thus also be considered as the number of production
steps (or, alternatively, the length of the value chain). By making management easier (thus
enhancing productivity of skilled labor), business services sustain more production steps,
thereby exploiting the productivity a finer division of labor within the value chain might
imply. Since we abstract from industrial-organization aspects, it makes no difference whether
the production steps are made in-house or outsourced and subsequently assembled. Instead,
the focus is on whether business services are produced locally or offshore. We will come
In order to avoid corner solutions, we impose a priori limits on productivity parameters such that 1 > α+2β.
back to this issue shortly. In principle, the same underlying production structure applies to
a representative firm j in Foreign. Costs of firms producing differentiated goods are thus
i) labor costs, wL Li + wH Hi , ii) fixed (set-up) costs as measured by X i , and iii) costs for
business services zi which presumably can be purchased at price γ (or have opportunity costs
of γ). The profit function of a representative firm i in Home’s differentiated good sector is
πi = Ai pi − wL Li − wH Hi − γzi − X i
As regards business services, we suppose that they are supplied by m firms (with the
number of firms, however, endogenously determined) and produced with high-skilled labor
according to the following production function
B ≡ b (Hn+1 )δ
with productivity parameters δ < 1 and b > 0. Parameter δ is supposed to capture advances
in ICT in particular. The higher δ, the higher productivity. Advances in ICT will provide
one of the drivers of trade (as well as compensation and employment) in business services.
As previously adumbrated, we assume here that in addition to labor costs, wH Hn+1 , the m
firms share in the costs of a joint input (think, for instance, of the financing of a university),
X n+1 , which is supplied by the numéraire sector. Hence, each firm incurs fixed costs, X n+1 /m,
independent of it’s own scale of production. However, the more firms, the lower the individual
contributions. These properties will beef-up the productivity-enhancing effect of service links
in the production of differentiated goods as trade integration widens. If business services are
sold at price γ (which however, is endogenously determined by supply and demand), profit
functions in business services are
πn+1 = γb (Hn+1 )δ − wH Hn+1 −
X n+1
3. Closing the Model
With free entry at the local level average costs equal price. In addition, we assume that
differentiated goods as well as business services are internationally tradable while x0 is not.
However, due to specialization in differentiated goods and business services accounting for
factor price equalization, the numéraire-good carries the same price in Home and in Foreign,
although it is itself considered non-tradable. Since we restrict our attention to situations
for which factor price equalization obtains via trade, profit maximization in the numéraire
sector fixes low-skilled wages at wL = a, both, at home and abroad.
Corresponding first order conditions in the differentiated-goods sector imply industryspecific demand for high- and low-skilled labor as well as business services of a representative
firm i in Home according to
Li =
αAi pi
Hi =
βAi pi
and zi =
βAi pi
In Home’s business-services production, profit maximization yields demand for high-skilled
labor of a representative firm
Hn+1 =
With perfect labor mobility across sectors (and firms) within each segment (high and low
skilled), intersectoral allocation is driven by equality of real wages and marginal productivity.
Qua symmetry assumption, the same applies to Foreign. With factor price equalization via
specialization in differentiated goods resp. business services we thus obtain the following set
of equilibrium conditions
wL = wL∗ = a,
wH = wH
and γ = γ ∗ .
for the two sorts of labor and business services at price γ, γ ∗ equalized via arbitrage.
Free entry into all businesses results in the following zero-profit conditions
γb (Hn+1 )δ − wH Hn+1 −
Ai pi − wL Li − wH Hi − γzi − X i = 0;
X n+1
= 0.
Finally, all factor and product markets must be cleared. Local factor market equilibrium
requires that supply for each category of labor equals demand
L = L0 + nLi
H = mHn+1 + nHi
L = L∗0 + n∗ L∗i
H = m∗ Hn+1
+ n∗ Hi∗
Since the numéraire is considered non-tradable, its markets must be cleared locally as well
a L − nLi = µb wL L + wH H + nX i + X n+1
a L − n∗ L∗i = µb wL∗ L + wH
+ n∗ X i + X n+1
World output of a representative differentiated good i, by contrast, must equal world demand
Ai = ci + c∗i
and analogously for any differentiated good j supplied by a Foreign firm. Due to trade, the
same applies to business services, i.e.
mB + m∗ B ∗ =
z +
i=1 i
j=1 j
In equilibrium, then, international trade must be balanced, that is
i=1 i
− mB =
Ai pi −
p i ci +
p∗ c˜
j=1 j j
Note though that Walras’ law implies that, if local markets for the numéraire clear, trade is
balanced as well – and vice versa.13
This completes our model of business-services offshoring with a particular focus on the
role of service links in the management of value chains.14 Exogenous variables are thus the
productivity parameters (α, β, δ, a and b), as are the fixed input requirements (X i = X j ,
X n+1 = X n+1 ). All other variables are endogenously determined, that is, the number of firms
in the differentiated-goods and the business-services sector, (n, n∗ ) and (m, m∗ ) respectively,
consumption (c0 , ci , c˜j ; c∗0 , c∗i , c˜j ∗ ) and therefore output (Ai , A∗j and B, B ∗ ), business-services
inputs (zi , zj∗ ), as well as the relative price of business services and of differentiated goods (γ,
pi , p∗j ). Furthermore, we have as endogenous variables employment in the numéraire sector
and in differentiated goods production (L0 , Li , L∗j , Hi , Hj∗ ), and, finally, in business services
(Hn+1 , Hn+1
1. Advances in ICT: Resolving the Complementarity Puzzle
First consider the impact of advances in ICT as captured by parameter δ on business service
offshoring, skilled labor, and occupations. Table 2 summarizes results. Although focusing
on the effects from Home’s perspective, implications for Foreign can be easily derived, with
the results basically symmetrical.
< insert Table 2 about here >
Recall that, indirectly, advances in ICT also foster productivity in differentiated-goods
production which proves to be crucial for the impact on skilled labor in the offshoring country.
Most importantly, we obtain: a shift towards business services occupations, notably, both on
a national as well as on a world-wide account. Hence, business-services offshoring does not
An Appendix containing a proof (in section A1) is available upon request from the authors.
Section A2 in the Appendix (see previous footnote) compiles the details in solving the model.
necessarily come at the expense of the skilled working in these occupations in the country
that produces more of these services offshore due to ICT, that is, with the trade in business
services to GNP ratio rising.15 At the same time, the skill premium increases. As a result,
the share of business services in total value added increases as well, as does the share of
business services in trade. Notably, these results apply to both of the countries, Home and
Foreign, and are thus comprehensive while traditional approaches usually differ according to
(relative) factor proportions. Moreover, the traditional North-South perspective maintains
that business-services offshoring comes at the expense of skilled labor via a pressure on
wages. Not so in our case. Rather, there is an element of complementarity between business
services offshoring and labor market prospects of the skilled, even in those occupations that
are seemingly threatened by offshoring. What essentially drives these results is the economic
characteristics in the production of business services in conjunction with the producitivityenhancing effects of service links with the former providing for a leverage effect on the latter.
This process is key to explaining the stylized facts in empirical studies that are puzzling and
that are on face of them difficult to reconcile, namely that skilled labor does better despite
of more service offshoring.
In Foreign, the impact of advances in ICT are symmetrical. The difference is mostly in the
direction of trade flows and in the denominator, i.e. H versus H , which, however, is invariant
to changes in productivity within the business-services industry. Consequently, the share of
business-services employment in total employment increases, independent of location. The
same holds for the share of value added of business services in GNP. Since the numerator is
in any case the same, differences in the ratio in Home and in Foreign are again solely due to
differences in the denominator, i.e., in this case, GNP (Y , Y ∗ ). However, because of factor
price equalization, wages are the same, as are – qua assumption – factor proportions, so that
differences here, as well, reduce to the scaling factor in the numerator, H versus H .
Trade to GNP ratios increase in Home and in Foreign: since Home’s imports are Foreign’s exports, trade
flows are the same, except for the sign. Hence, in terms of absolute values, the numerator is in any case
the same, while differences lie in the denominator only, which boils down to H versus H .
2. Widening of Trade Integration: The Not So Bright Outlook for the Skilled
Next, consider a widening of trade integration, modeled as an increase of the size of For∗
eign as measured by endowments H . Table 3 summarizes the main results with each row
corresponding to the respective row and variable in Table 2.
< insert Table 3 about here >
Notably, while advances in ICT come to the benefit of skilled labor employed in business
services even if those services are internationally tradable and even if, as a consequence,
offshoring increases, the outlook is not quite as bright when it comes to trade integration.
For in this case the competitive element is much stronger to the effect that it might well
dominate vis-à-vis any productivity impact. While nevertheless trade in general increases
relative to GNP, the share of trade in business services heads down, both when compared
to GNP and other trade flows. What happens is onshoring, with Home’s employment in
business services as well as their share in total value added increasing. However, the flip side
of this seemingly “positive” development is a decline in the skill premium. Hence, onshoring
is not necessarily a sign for skilled labor in business services doing better. Or to put it
differently: skilled business services should probably cheer offshoring instead of fearing it.
Whether more services are produced onshore or offshore is not necessarily a good indicator
for concerns when it comes to skilled labor in offshorable business services. What matters is
what drives these results, with advances in ICT and the widening of trade integration tending
to work in opposite directions. Note that this also applies to Foreign. Here, as well, trade
integration tends to curb trade in business services, however, at the cost of lowering the skill
premium. Advances in ICT, by contrast, tend to push not only trade in general, but, also
both, trade in business services and the skill premium. Henceforth, policy is well advised not
to glare exclusively at a single indicator nor to generally condemn offshoring. Rather, based
on these results, it should embrace advances in ICT and provide a flourishing environment
for start ups in these branches, even though they might eventually foster service offshoring.
The increased tradability of business services raises concerns that skilled labor might be the
next “victim” of globalization in the sense that firms substitute expensive domestic labor for
cheap imports of those services. While some empirical studies seemed to cater to this notion,
others found traces of a puzzling co-movement between offshoring and certain segments of
the labor market resp. employment at the firm level, both of which proved difficult to
reconcile with the traditional substitution-led perspective on offshoring. An explanation of
the phenomenon is still pending, as is an examination of whether the amount of offshoring is
indeed a reliable indicator for how the skilled fare in an increasingly globalized environment.
In order to shed more light on this issue, we present a novel theoretical framework that i)
shows that there are two major engines in services offshoring and that ii) allows to disentangle
the transmission channels and the impacts of the two of them. This tool kit reveals that
they are associated with very different results. While both of them affect production and
specialization at the intensive as well as the extensive margin, some of their effects on the
skilled actually tend to work in opposite directions. Hence, net impacts depend on which of
them prevails. While trade integration per se indeed tends to be associated with a depressing
effect on skill premia, advances in ICT tend to come to the benefit of skilled labor in business
services – notably, even if accompanied by more imports of those services.
Hence, the fear of service offshoring is not necessarily warranted and the outcome does
not quite fit neatly into the widely held perception that offshoring comes at the expense
of the corresponding labor market segments at home. To some extent, these results may
remind oneself of the trade versus technology discussion in the widening of the wage gap and
material offshoring. In this way, Kletzer (2009b, 327) can be interpreted in noting that “In
this sense, we are back to the trade versus technology debate, last seen in the many papers
on international competition and the decline of manufacturing employment”. However, with
an important twist: apart from the fact that the channels of material and services offshoring
are very different, business services have an additional productivity-enhancing effect that is
actually beefed up by advances in ICT and therefore something to cheer about. It is this
leverage effect that makes for a substantial difference. At any rate, gauging the consequences
by focusing on the amount of business-services offshoring (and thus trade) can be quite
misleading – and inappropriate, if not counterproductive, when it comes to policy responses.
Abramovsky, L.; Griffith, R. 2006. Outsourcing and Offshoring of Business Services: How Important is ICT? Journal of the European Economic Association 4(2-3), 594-601.
Amiti, M.; Wei, S.-J. 2005. Fear of Service Outsourcing: Is it Justified? Economic Policy 20(42),
Amiti, M.; Wei, S.-J. 2009. Does Service Offshoring Lead to Job Losses? Evidence from the
United States. In Reinsdorf, M.; Slaughter, M.J. (eds.), International Trade in Services and
Intangibles in the Era of Globalization. Chicago: University of Chicago Press, 227-43.
Antràs, P. forthcoming. Grossman-Hart (1986) Goes Global: Incomplete Contracts, Property
Rights, and the International Organization of Production. Journal of Law, Economics and
Arndt, S.W. 1997. Globalization and the Open Economy. North American Journal of Economics
and Finance 8(1), 71-79.
Arndt, S.W. 1998. Super-specialization and the Gains from Trade. Contemporary Economic
Policy XVI, 480-85.
Aspray, W. 2010. IT Offshoring and American Labor. American Behavioral Scientist 53(7),
Baldwin, R. 2009. How Many Jobs are Onshorable? Re-interpreting the Blinder Numbers in the
Light of New Trade Theory. Posted on VoxEU, June 15, 2009 (http://www.voxeu.org/article/
how-many-jobs-are-onshorable, retrieved Nov. 03, 2012).
Baldwin, R.; Robert-Nicoud, F. 2011. Trade-in-goods and Trade-in-tasks: An Integrating Framework. University of Geneva Working Paper Services WPS 11112, Nov.
Blinder, A.S. 2006. Offshoring: The Next Industrial Revolution? Foreign Affairs, March/April,
Blinder, A.S. 2009. How Many U.S. Jobs Might Be Offshorable. World Economics 10(2), 41-78.
Blinder, A.S.; Krueger, A.B. forthcoming. Alternative Measures of Offshorability: A Survey
Approach. Journal of Labor Economics. Previous version: NBER Working Papers 15287,
Aug. 2009.
Crinò, R. 2009. Offshoring, Multinationals and the Labour Market: A Review of the Empirical
Literature. Journal of Economic Surveys 23(2), 197-249.
Crinò, R. 2010. Service Offshoring and White-Collar Employment. The Review of Economic
Studies 77(2), 595-632.
Crinò, R. 2012. Service Offshoring and the Skill Composition of Labour Demand. Oxford Bulletin
of Economics and Statistics 74(1), 20-57.
Deardorff, A.V. 2001a. Fragmentation in Simple Trade Models. The North American Journal of
Economics and Finance 12(2), 121-37.
Deardorff, A.V. 2001b. International Provision of Trade Services, Trade, and Fragmentation.
Review of International Economics 9(2), 233-48.
Deardorff, A.V. (2005): A Trade Theorist’s Take on Skilled Labor Outsourcing. International
Review of Economics & Finance 14(3), 259-71.
Debaere, P.; Görg, H.; Raff, H. forthcoming. Greasing the Wheels of International Commerce:
How Services Facilitate Firms’ International Sourcing.Canadian Journal of Economics.
Desmet, K.; Ghani, E.; O’Connell, S.D.; Rossi-Hansberg, E. 2012. The Spatial Development of
India. The World Bank Policy Research Working Paper 6060, May.
Dluhosch, B. 2006. Intraindustry Trade and the Gains from Fragmentation. North American
Journal of Economics and Finance 17(1), 49-64.
Dluhosch, B. 2010. ICT: Inflating or Deflating Offshoring? Journal of International and Global
Economic Studies 3(1), 1-24.
Escaith, H. 2008. Measuring Trade in Value Added in the New Industrial Economy: Statistical
Implications. MPRA Paper 14454, University Library of Munich.
Feenstra, R.C.; Hanson, G.H. 2003. Global Production Sharing and Rising Inequality: A Survey of
Trade and Wages. In Choi, K.; Harrigan, J. (eds.) Handbook of International Trade. Oxford:
Basil Blackwell, 146-87.
Francois, J.F. 1990. Trade in Producer Services and Returns Due to Specialization Under Monopolistic Competition. Canadian Journal of Economics 23(1), 109-24.
Francois, J.F.; Hoekman, B. 2010. Services Trade and Policy. Journal of Economic Literature
48(3): 642-92.
Francois, J.F.; Woerz, J. 2008. Producer Services, Manufacturing Linkages, and Trade. Journal
of Industry, Competition and Trade 8(3), 199-229.
Freund, C.; Weinhold, D. 2002. The Internet and International Trade in Services. American
Economic Review (Papers and Proceedings) 92(2), 236-40.
Friedman, T.L. 2007. The World is Flat 3.0. New York: Picador.
Geishecker, I.; Görg, H. forthcoming. Services Offshoring and Wages: Evidence From Micro Data.
Oxford Economic Papers.
Ghani, E. (ed.) 2010. The Service Revolution in South Asia. Oxford: Oxford University Press.
Gonzales, F.; Jensen, J.B.; Kim, Y.; Kyvik Nordås, H. 2012. Globalisation of Services and Jobs. In
Lippoldt, D. (ed.) Policy Priorities for International Trade and Jobs. Paris: OECD, 175-92.
Grossman, G.M.; Rossi-Hansberg, E. 2008. Trading Tasks: A Simple Theory of Offshoring. American Economics Review 98(5): 1978-97.
Grossman, G.M.; Rossi-Hansberg, E. 2012. Task Trade Between Similar Countries. Econometrica
80(2), 593-629.
Harrison, A.E.; McMillan, M.S.; Null, C. 2007. U.S. Multinational Activity Abroad and U.S. Jobs:
Substitutes or Complements? Industrial Relations 46(2), 347-65.
Hijzen, A.; Pisu, M.; Upward, R.; Wright, P.W. 2011. Employment, Job Turnover, and Trade in
Producer Services: UK Firm-level Evidence. Canadian Journal of Economics 44(3), 1020-43.
Horgos, D. 2011. International Outsourcing and the Sector Bias: New Empirical Evidence. Review
of International Economics 19(2), 232-44.
Hummels, D.; Ishii, J.; Yi, K.-M. 2001. The Nature and Growth of Vertical Specialization in
World Trade. Journal of International Economics 54, 75-96.
Jensen, J.B. 2011. Global Trade in Services: Fear, Facts, and Offshoring. Washington D.C.:
Peterson Institute for International Economics.
Jensen, J. B.; Kletzer, L.G. 2006. Tradable Services: Understanding the Scope and Impact of
Services Offshoring. In Collins, S.M.; Brainard, L. (eds.) Brookings Trade Forum 2005:
Offshoring White-Collar Work. Washington, D.C.: Brookings Institution, 75-134.
Jensen, J. B.; Kletzer, L.G. 2010. Measuring Tradable Services and the Task Content of Offshorable Services Jobs. In Abraham, K.G.; Spletzer, J.R.; Harper, M. (eds), Labor in the
New Economy. NBER Book, Chicago: University of Chicago Press, 309-35.
Jones, R.W.; Kierzkowski, H. 1990. The Role of Services in Production and International Trade:
A Theoretical Framework. In Jones, R.W.; Krueger, A.O. (eds.) The Political Economy of
International Trade. Oxford: Basil Blackwell, 31-48.
Jones, R.W.; Kierzkowski, H.; Lurong, C. 2005. What Does Evidence Tell Us About Fragmentation
and Outsourcing? International Review of Economics & Finance 14, 305-16.
Kirkegaard, J.F. 2008. Offshoring, Outsourcing and Production Relocations – Labor Market
Effects in the OECD and Developing Asia. The Singapore Economic Review 53(3), 371-418.
Kletzer, L.G. 2009a. Understanding the Domestic Labor Market Impact of Offshore Services
Outsourcing: Measurement Issues. Paper prepared for Measurement Issues Arising from
the Growth of Globalization, Conference sponsored by the W.E. Upjohn Institute for Employment Research and the National Academy of Public Administration, Nov. 6-7, 2009
(http://www.upjohninst.org/measurement/final-program.html, retrieved Nov. 12, 2012).
Kletzer, L.G. 2009b. Comment on van Welsum and Reif, We Can Work It Out: The Globalization
of ICT-Enabled Services. In Reinsdorf, M.; Slaughter, M.J. (eds) International Trade in
Services and Intangibles in the Era of Globalization. Chicago: University of Chicago Press,
Kohler, W.; Wrona, J. 2011. Offshoring Tasks, Yet Creating Jobs? University of Tübingen Working
Papers in Economics and Finance No. 12.
Levy, F.; Murnane, R.J. 2005. The New Division of Labor. Princeton: Princeton University Press.
Lewin, A. (2011): Trade in Services: The Global Sourcing of Business Services. In Ramamurti, R.;
Hashai, N. (eds), The Future of Foreign Direct Investment and the Multinational Enterprise
(Research in Global Strategic Management, Vol. 15). Bingley: Emerald Publishing Group,
Lipsey, R.E. 2009. Measuring International Trade in Services. In Reinsdorf, M.; Slaughter, M.J.
(eds) International Trade in Services and Intangibles in the Era of Globalization. Chicago:
University of Chicago Press, 27-70.
Liu, R.; Trefler, D. 2008. Much Ado About Nothing: American Jobs and the Rise of Service
Outsourcing to China and India. NBER Working Paper 14061, Cambridge/Mass., June.
Liu, R.; Trefler, D. 2011. A Sorted Tale of Globalization: White Collar Jobs and the Rise of
Service Offshoring. NBER Working Paper 17559, Cambridge/Mass., Nov.
OECD 2012. Internet Economy Outlook, Paris: OECD.
Spencer, B.J. 2005. International Outsourcing and Incomplete Contracts. Canadian Journal of
Economics 38, 1107-35.
Sturgeon, T.J.; Nielsen, P.B.; Linden, G.; Gereffi, G.; Brown, C. (2012): Direct Measurement
of Global Value Chains: Collecting Product- and Firm-Level Statistics on Value Added and
Business Function Outsourcing and Offshoring. Mimeo: MIT. Forthcoming as Ch. 9 in the
World Bank Volume: The Fragmentation of Global Production and Trade in Value-Added Developing New Measures of Cross Border Trade.
Trefler, D. 2005. Services Offshoring: Threats and Opportunities. In Brookings Trade Forum:
Offshoring White-Collar Work. Washington, D.C.: Brookings Institution, 35-73 (including
Discussion by D. Rodrik and P. Antraàs).
UNCTAD 2004. World Investment Report 2004: The Shift Towards Services. New York and
Geneva: United Nations Conference on Trade and Development.
Van Welsum, D.; Vickery, G. 2005. Potential Offshoring of ICT-Intensive Using Occupations.
Report to the Working Party on the Information Economy (WPIE), OECD Document DSTI/
ICCP/IE(2004)19/FINAL, Paris: OECD.
WTO 2012a. World Trade Report, Geneva: WTO.
WTO 2012b.
International Trade Statistics, Geneva:
WTO (http://www.wto.org/
english/res_e/statis_e/its2012_e/its12_toc_e.htm; data retrieved Nov. 2, 2012).
Yi, K.-M. 2003. Can Vertical Specialization Explain the Growth of World Trade? Journal of
Political Economy 111(1), 52-102.
Yu, K.-H.; Levy, F. 2010. Offshoring Professional Services: Institutions and Professional Control.
British Journal of Industrial Relations 48, 758-83.
high skilled
low skilled
inputs to
goods sector
services sector
goods sector
i=1 zi
X n+1
inputs to
i=1 ci
Table 1. Input-Output Matrix: Home
i=1 zi
i=1 ci
c0 +
X i + X n+1
∗ ∗
i −n p ci
trade in business services
to GNP ratio
(national level)
trade to GNP ratio
share of business
services in trade
∗ ∗
γ(nzi −mB)+npc∗
i +n p ci
γ(nzi −mB)
(1 − µ̂)
+2(1+δ)X n+1 H
γ(nzi −mB)
nAp−γ(nzi −mB)
(1−δ) δaκ H+H
γ(nzi −mB)
δ 2 X n+1 H+H
δ H+H
(1−δ) H+H
(1+δ)2X n+1
share of business services
in total value added
(national level)
share of business
services employment
in total skilled employment
share of business
in skilled employment
at national level
skill premium
wH mHn+1
δ+4X n+1
+2(1+δ)X n+1
(1−µ)µ2(1−α−2β)X n+1
2 > 0 since H > H ∗
H −H
2H(1+δ)−δ H+H
2 H+H
((1−µ)(n+n∗ )+µ)+2n
((1−µ)(n+n∗ )+µ)2 (1−δ)2 βXi H+H
Note though that what matters is absolute values of imports
and exports, which means that the sign of the first derivative
with respect to δ is actually positive for all H , H . Moreover,
since trade is balanced and Foreign the mirror image of Home,
the same applies to Foreign and, thus, also to world trade.
− 21 H∗
∗ (1−α−2β)X n+1 4H− H+H
to GNP ratio increases in δ
((1−µ)(n+n∗ )+µ)2 (1−δ)2 βXi
trade in business services
inspection of (??) shows that the 2nd term
increases in δ, as does the 1st term
aκ H+H
(1−δ)2 δaκ H+H
δX n+1 H+H
2(1+δ)2 H
4X n+1
Derivative with respect to δ
Table 2. Advances in ICT: impact on services employment and compensation, value added and trade
share of business
services employment
in total skilled employment
share of business
services in trade
trade to GNP ratio
trade in business services
to GNP ratio
(national level)
share of business services
in total value added
(national level)
share of business
in skilled employment
at national level
2 < 0
∗ 2
2 < 0
aκH +
2(1+δ)H−δ H+H
+(1−δ)aκ H+H
2(1+δ)X n+1
(1−δ) H+H
2(1+δ)X n+1 4HH +δ H−H
δ ∧ H > H / (2 + δ) a sufficient condition
2µ̂(1−α−2β)(1−δ) H+H
β(1+δ)X i
−β(1+δ) H−H
2(1+δ)H−δ H+H
− (1−µ)
+2(1+δ)X n+1 H
(1−µ) 4(1−α−2β)X n+1 (1+δ)H
∗ 2
β(1−δ)X i H H+H
− (1 − µ̂)
(1−δ) δaκ H+H
δ 2 X n+1 H+H
(1−δ) H+H
(1+δ)2X n+1
Derivative with respect to H
skill premium
(see Table 1 for
corresponding variables)
+δ H−H
Table 3. Widening of trade integration: impact on services employment and compensation, value added and trade
Fly UP