Journal of International Business Studies (2016) 47, 563–576

© 2016 Academy of International Business All rights reserved 0047-2506

www.jibs.net

The impact of information and communication

technology adoption on multinational firm
boundary decisions

Wenjie Chen1 and Abstract

This article evaluates the effect of adopting Internet-enabled information and
Fariha Kamal2 communication technology (ICT adoption) on the decision to reorganize
1
production across national borders (foreign boundary decision) by multinational
Department of International Business, George
enterprises (MNEs). Using a transaction cost framework, we argue that ICT
Washington University, Washington DC, USA;
2 adoption influences foreign boundary decisions by lowering coordination costs
Center for Economic Studies, US Census Bureau,
Suitland, MD, USA both internally and externally for the firm. We propose that the heterogeneity in
the technology’s characteristics, namely complexity and the production pro-
Correspondence: cesses’ degree of codifiability, moderate this influence. Using a difference-in-
W Chen, Department of International differences methodology and exploiting the richness of confidential US Census
Business, George Washington University, Bureau microdata, we find that overall ICT adoption is positively associated with
Washington DC 20052, USA. greater likelihood of in-house production, as measured by increases in intra-firm
Tel: +1 (202) 994 3275 trade shares. Furthermore, we find that more complex forms of ICT are
associated with larger increases in intra-firm trade shares. Finally, our results
indicate that MNEs in industries in which production specifications are more
easily codified in an electronic format are less likely to engage in intra-firm
relative to arm’s length trade following ICT adoption.
Journal of International Business Studies (2016) 47, 563–576. doi:10.1057/jibs.2016.6

Keywords: transaction cost theory; information technology; intra-firm trade; intra-

organizational structures; IT/IS infrastructure

INTRODUCTION
Profound innovations in information and communication technol-
ogies (ICTs) during the last three decades have reshaped the global
landscape in which firms operate. The emergence of these technol-
ogies has enabled firms to reduce transactions costs, including those
in transportation, communication, and coordination. Multinational
enterprises (MNEs), in particular, have benefited from the innova-
tions in ICT and have adopted them into their operations and
international production processes. This in turn has influenced
increasingly complex firm boundary decisions.
The transaction cost theory has been the leading paradigm in
the analysis of the multinational firm, particularly, to answer the
Received: 11 January 2014
question of whether or not to undertake an activity within a firm’s
Revised: 15 December 2015
Accepted: 29 December 2015 own boundaries (Buckley & Casson, 1976, 1998, 2009; Dunning,
Online publication date: 3 March 2016 1980; Hennart, 1982, 2009; Teece, 1986). Under this conceptual
 The impact of ICT adoption Wenjie Chen and Fariha Kamal
564

framework firms choose over alternative govern- evidence on the contrary to suggest that, on balance,
ance structures in order to minimize transaction increased ICT investments are positively correlated
costs. By using ICT, MNEs are able to mitigate with intra-organizational interactions. Both these
transaction costs that might alter the decision studies capture the net effect of ICT investments on
between in-house production vs arm’s length foreign boundary decisions, with the underlying
exchange. Forman and McElheran (2012) study premise that outsourcing and in-house production
the effect of ICT adoption on the domestic bound- are substitutes for each other.
ary choices of firms by examining changes in the However, ICT affects all MNE-related production,
share of total shipments that are transferred both in-house and at arm’s length, in ways that are
internally within domestic firms in the United not always mutually exclusive. In a departure from
States. In contrast, Fort (2013) primarily focuses prior research, our study highlights that the choices
on the effect of ICT on domestic vs foreign sour- of arm’s length and intra-firm production may occur
cing decisions, thus highlighting the spatial loca- simultaneously within the same firm. For instance,
tion choice of the firm. Our study complements in early 2000, the Swiss food and beverage MNE
these existing studies by analyzing the role of ICT Nestlé invested US$1.8 billion in ICT (Worthen,
adoption in shaping the decision between how 2002).1 After ICT installation, global subsidiaries
much is produced within and outside the bound- exploited the new network to share price informa-
aries of the firm across national borders, hence- tion and exert market power by acting in a coordi-
forth referred to as a firm’s foreign boundary nated manner. As a result, Nestlé reduced the
decision. number of external suppliers and procurement costs
We examine ICT’s impact on a firm’s foreign by up to 20%. Concurrently, Nestlé began using a
boundary decision to make or buy through the lens new web-based system to contract with external
of the firm’s intra- and inter-firm trade choices, suppliers directly. ICT installation can thus have
explicitly assuming that the firm has decided on significant effects on firms’ boundary choices and
foreign production. We propose that from the view- as a result, its intra-firm and arm’s length trading
point of the firm, ICT adoption lowers both internal patterns.
and external coordination costs ex post through two Many studies in the organization and communica-
independent effects: (1) lowering of communication tion science research literature have examined the
and search costs (Malone, Yates, & Benjamin, 1987), relationship between ICT and organizational struc-
and (2) lowering of economic incentive costs in tures (McPhee & Poole, 2001; Rice & Gattiker, 2001),
terms of economic moral hazards and opportunistic but relatively little research attention has focused on
behavior (Clemens, Reddi, & Row, 1993; Gurbaxani the specificity and heterogeneity of ICT systems and
& Whang, 1991). The efficient mix of a firm’s in- their distinct moderating effects on a firm’s bound-
house and arm’s length production is determined by ary decisions (notable exceptions are Feinberg &
trading off external and internal coordination costs. Keane, 2006; Keane & Feinberg, 2007; Rangan &
We argue that this baseline effect of ICT on intra- Sengul, 2009). In this article, we study the moderating
firm relative to arm’s length production is theoreti- role of characteristics of ICT systems. We hypothe-
cally ambiguous and hence ultimately an empirical size that the firm’s intra-firm transactions relative
question. to its arm’s length transactions will increase, the
Most existing studies – with the exception of greater the complexity of the type of ICT adopted.
Feinberg and Keane (2006) and Keane and Feinberg More complex types of technology require greater
(2007) – in the area of ICT and firm organization synchronization between adopting firm units in
have generally concentrated on the net effect of order to maximize the gains. Thus ex post of the
the impacts of ICT adoption on firm boundary firm’s decision to adopt ICT, it is more likely that
decision, that is, either increased in-house produc- more complex ICT would require using parties to
tion or increased arm’s length production. De la have the necessary ICT infrastructure and commit-
Torre & Moxon (2001) provide an excellent over- ment in investing and maintaining the network
view of the literature pertaining to the same time systems.
period, the early 1990s, as the focus in our study. We further hypothesize that the firm’s intra-firm
Rangan and Sengul (2009) find that MNEs in indus- transactions relative to its arm’s length transactions
tries with higher ICT investment intensity exhibit will decrease the more readily its production process
lower propensity for transnational integration. Dosi, can be electronically codified. As it is easier to break
Gambardella, Grazzi, and Orsenigo (2008) provide up a production chain in which processes are more

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
565

readily codified, each value chain link can separately intra-firm or arm’s length trade. We re-estimate our
benefit from ICT adoption and result in cost reduc- baseline specification controlling for a firm’s prior
tions. Thus we would expect MNEs operating in use of proprietary network technology.
industries where the production process is more Our examination of the role of ICT on a firm’s
readily codified into electronic formats to have a foreign boundary decisions makes a theoretical and
proclivity for arm’s length relative to intra-firm empirical contribution to the existing literature
transactions following ICT adoption. on ICT and firm organizational choices. Although
We study ICT adoption in the context of globaliza- there is a growing literature on the role of frag-
tion and increased trade integration (Antràs & mentation and offshore production, where techno-
Yeaple, 2014). Our sample period (1992 and 1999) logical improvements play an essential role (Antràs,
spans the advent of Internet-enabled ICT and Rossi-Hansberg, & Garicano, 2008; Costinot, Vogel,
the emergence of global value chains. As Internet- & Wang, 2013; Deardorff, 2001; Grossman &
enabled ICT has been commercially available only Rossi-Hansberg, 2008; Jones & Kierzkowski, 2001;
since 1995 (Forman, Goldfarb, & Greenstein, 2005), Kohler, 2004), the theoretical and empirical analyses
we focus on a relatively early period of Internet- rely on the assumption that firms have already
enabled ICT adoption. Thus our analysis provides a upgraded their technology in order to facilitate pro-
valuable benchmark that can act as a baseline and duction fragmentation. Our theoretical and empirical
lower bound on the impact of ICT adoption on focus is exactly on the effect of a firm’s decision to
MNEs’ boundary decisions. We use detailed data adopt ICT on its foreign boundary choices. We build
on US manufacturers’ choice of ICT adoption at the on the work of Rangan and Sengul (2009) by extend-
plant level for a host of different industries in ing the theoretical model to include more detailed
the manufacturing sector using the 1999 Computer types of ICT applications and the moderating effects
Network Use Supplement (CNUS), an addendum to of characteristics of ICT on a firm’s international
the Annual Survey of Manufactures (ASM). This boundary decisions. Our work also relates to and
novel and comprehensive data set allows us to extends the study by Feinberg and Keane (2006),
provide a nuanced picture of the role of ICT adop- which finds that technology, measured as the residual
tion on MNE boundary decisions and unlike most from a structural model, is an important factor in
existing studies, we are able to exploit the hetero- intra-firm trade by multinationals between the Uni-
geneity of the disaggregated data in terms of various ted States and Canada. Whereas Rangan and Sengul
types of ICT adoption. We measure MNE’s foreign (2009) and Cho (1990) employ data aggregated at the
boundary choices through its intra-firm and arm’s industry level, our analysis uses a direct plant-level
length trade transactions obtained from the Long- measure of ICT and provides evidence of a link
itudinal Foreign Trade Transactions Database (LFTTD). between specific types of ICT adoption and a firm’s
We find robust empirical support for our hypo- international boundary decision.
theses. Overall, ICT adoption enhances within-firm The advent of ICT has been accompanied by
boundary transactions as measured by increased increased formation of global value chains that are
intra-firm trade shares. By using a difference-in- creating new opportunities for firms to become more
differences identification strategy and contrasting globally integrated, which has been associated with
pre- and post-ICT adoption periods, results show growth (Baldwin & Lopez-Gonzalez, 2014). Thus it is
that firms with more sophisticated types of ICT important for policymakers to understand how new
experience increases in intra-firm relative to arm’s avenues linking them to global production chains
length trade transactions. We also confirm our can be exploited. Because MNEs are a central feature
expectation that following the adoption of ICT, of global production networks mediating 80% of
firms in industries where production processes are global trade (UNCTAD, 2013), understanding how
more readily codified into electronic formats experi- ICT reshapes multinational firms’ global activities is
ence decreases in intra-firm trade shares. Our base- an important avenue of scholarly investigation.
line results are robust to construction of alternative The following section develops our hypotheses
samples and controls for simultaneity bias. We con- grounded in transaction cost theory. The third
sider two distinct samples – single-plant firms only section describes the data sources and variable
and single-plant and multi-plant firms where all its construction, the fourth section describes our
plants either adopt or do not adopt ICT. We address empirical methodology, the fifth section presents
the simultaneity bias that arises if firms are pro- the results, and the final section provides discus-
actively adopting ICT in anticipation of enhancing sion and conclusion.

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
566

HYPOTHESES DEVELOPMENT and provide empirical evidence for increased inter-

nationalization at arm’s length associated with
Transaction Costs and ICT Adoption increased ICT spending. Hitt (1999) observes that
An MNE has to communicate with, coordinate, an increased use of IT systems is associated with a
search for and monitor cross-border operations. substantial decrease in vertical integration, and
Broadly speaking, a firm’s coordination costs can suggests that the empirical evidence points to the
be divided into two distinct categories: internal effect of IT systems lowering external coordination
coordination costs and external coordination costs costs more than internal coordination costs. Malone
(Kim & Mahoney, 2005). Internal coordination et al. (1987) propose that ICT can lead to an overall
costs represent the economic costs incurred for com- decline in coordination costs in terms of searching
munications, data transfer, and other economic and communicating with external transacting
expenditures on managing dependencies between parties. Both Gurbaxani and Whang (1991) and
activities within a firm. External coordination costs Clemens et al. (1993) argue that the usage of ICT
represent the economic costs of locating suppliers, can reduce transaction risks such as the contractual
writing contracts, and other economic costs of mar- hazards of shirking and opportunism through
ket procurement. improved monitoring and reduced specificity in
The internalization literature has highlighted the coordination with external partners. Thus ICT sys-
opposing nature of the outsource/offshore vs produ- tems can reduce external coordination costs and
cing in-house decision yielding divergent findings. encourage outsourcing.
On the one hand, the adoption of ICT can reduce Given these simultaneous channels whereby ICT
the economic costs of internal coordination within can reduce costs for both external and internal
the firm by improving the quality and speed of transactions, Gurbaxani and Whang (1991) aptly
information processing and management’s decision note that “the net effect of IT on firm size varies
making. Moreover, by adopting ICT, management from situation to situation, depending on the cost
can reduce agency costs through improved monitor- structure of the firm and the modes of synergy
ing and evaluation systems. Bloom, Garicano, generated by integration” (71). Hence the net effects
Sadun, and Van Reenen (2014) find that adoption are theoretically ambiguous and we rely on empirical
of ICT within the same organization can reduce evidence to guide us.
communication costs as well as knowledge acquisi- Thus far, we have abstracted from heterogeneity in
tion costs. Moreover, Dosi et al. (2008) find that the types of ICT as well as differences across industries in
size distribution of firms and industrial concentra- the application of ICT. In the following sections, we
tion in the advanced economies over the past few develop two testable hypotheses. First, we address
decades have not changed despite the increasing the adoption of different types of ICT and how it
role of ICT. They deduce that intra-organizational moderates the effect of a firm’s choice in internal vs
interactions still dominate over market interactions, external forms of transactions. Second, we examine
contradicting the notion that there has been a signi- how exogenous industry characteristics influence
ficant movement toward market coordination due firm’s boundary choices through the use of ICT.
to ICT. Along the same lines, other studies have
also found greater internalization effects of ICT Complexity of ICT
(Brynjolfsson, 1994; Gurbaxani & Whang, 1991; ICT systems vary widely in terms of their complexity
Zenger & Hesterly, 1997). Thus ICT adoption and sophistication and with them, the level of
can reduce internal coordination costs and provide investment that is required of the firm to undertake.
management with the capability to manage an Complex types of ICT, entailing relation-specific
organization more effectively and facilitate increased investments and high sunk costs, often requires the
integration. firm to incur additional costs of coordination among
On the other hand, adoption of ICT can directly adopting firm units. We note here that a priori, it is
reduce market transaction costs by providing a cost- not clear what type of firm would be more prone
effective means to access and process market infor- to adopt complex types of ICT, and it is not the focus
mation, thus reducing external coordination costs of our study. Existing research shows that the
to the firm. ICT can also reduce market transac- costs to adopt business process innovations are vari-
tion costs by facilitating closer inter-firm links able depending on the mix between internal and
through information sharing and mutual monitor- external adjustment costs that are firm specific
ing. Rangan and Sengul (2009), for instance, argue (McElheran, 2015). She describes that the amount

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
567

of complementary adjustment needed to adapt to into electronic formats, and share production
new innovations involving “new processes, support- requirements and designs (Fort, 2013) with other
ing organizational structures and enabling ICT firms. For firms in these industries that are able to
was difficult to predict and varied widely by firm make use of CAD and CAE technologies to break up
and by the type of business process innovation their production into smaller and simpler tasks by
they were pursuing” (1200). Our argument is that in codifying product specifications, it is possible to
order to be effective ex post, however, the adoption minimize leakage of crucial proprietary knowledge
of firm-specific ICT requires matching invest- by choosing to outsource certain intermediate input
ments for co-specialization with other firm-specific production to external suppliers and maintain
resources, such as ICT-trained employees or ICT- in-house the production of intermediates that have
supported distribution network. Furthermore, ex post higher intellectual property content. From an exter-
of adoption, more complex types of ICT that have nal supplier’s viewpoint, the codifying process
features of enhancing coordination and information addresses hold-up problems by reducing the cost of
sharing would become less valuable if the partnering customization significantly, resulting in greater
units’ ICT systems are not compatible. diversification of the supplier’s risk and reducing
Even after adoption of ICT, more complex types costs of being a “hold-up hostage” to each buying
of ICT requires coordinated effort between transact- firm. Then, ICT adoption in industries where the
ing parties for optimizing the benefits. Moreover, production process is easily codified into electronic
complexity may lead to increased interdependence formats eases the outsourcing process by finding the
between transacting parties due to the need to lowest cost producer in the market place. We pro-
implement similar customization of technology. pose the following:
Thus given that firms have already adopted ICT, it is
Hypothesis 2: All else being equal, the higher
more likely that higher degree of complexity in the
the degree of electronic codifiability of the produc-
ICT is associated with more internal integration in
tion process of the MNE’s industry, the more likely
order to ensure maximization of benefits from ICT
it is to experience increased outsourcing following
adoption. Thus we hypothesize the following:
ICT adoption.
Hypothesis 1: All else being equal, the more
complex the type of ICT adopted, the more likely
it is to facilitate increased integration between DATA
internal firm units. Sample
We use four confidential micro data sets sourced
Electronic Codifiability of the Production Process from the US Census Bureau – ASM that includes the
Industry characteristics may influence the scope and CNUS, Census of Manufactures, LFTTD, and the
nature of the impacts of ICT adoption on a firm’s Longitudinal Business Database (LBD) – in order to
internal vs external integration. One of the main carry out the analysis. In what follows below, we
barriers to externalization is the non-codifiable nat- describe the steps to construct our analysis data set
ure of tasks and production specifications along with and then we describe our variables of interest in
specific customization requirements of the product. detail.
Industries in which the production process lends
itself to fragmentation would benefit from the adop- ICT data
tion of ICT in locating the lowest cost producers In order to measure ICT adoption within a firm,
for each chain of its various production stages. we utilize the 1999 survey addendum to the US
For certain industries, for instance, Sturgeon (2002) Census’s Annual Survey of Manufacturers called the
describes how production processes are broken CNUS. The CNUS is a unique data set due to its
down into distinct value chains, whereby linkages wealth of detail on the firm’s networking technolo-
are achieved between the chains via the transfer of gies at the plant level and its national representa-
codified information. tiveness as an addendum to a legally required Census
Technological innovations in manufacturing pro- survey. The survey was intended primarily to mea-
cesses, such as the introduction of computer-aided sure electronic commerce (e-commerce) over com-
design (CAD) and computer-aided engineering (CAE) puter-mediated networks (Messenbourg, 2001).
software, have made it possible for the firm to Our measure of ICT captures computer networking
translate manufacturing designs and specifications technologies that can be used to communicate with

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
568

and share information between both internal and between two time periods. Then we merge the
external parties, and is therefore ideally suited for resulting data set to the CNUS data. The CNUS data
our analyses. In particular, we use the first ques- is first aggregated at the firm level prior to the merge,
tion in the CNUS that asks if the responding estab- and we describe the details of the aggregation
lishment uses various types of computer networks: method below. We further restrict our analysis sam-
Internet, intranet, extranet, Local Area Network ple to firms that operate in the manufacturing sector,
(LAN), Electronic Data Interchange (EDI). The asso- since CNUS was sent to manufacturing establish-
ciated variables in the survey data are separate ments only. Using the LBD, which consists of data
indicator variables that take on a value 1 if the plant on all existing establishments that have at least one
uses any one of these types of networks, and 0 paid employee in the US non-farm, private economy
otherwise. We use the first four indicator variables (Jarmin & Miranda, 2002), we identify the primary
to create an overall ICT measure, described shortly. two-digit sector – manufacturing, retail, wholesale
We consider LAN, Internet, intranet, and extranet trade, or others – in which a firm operates. Since
as Internet-enabled network technologies since multi-unit firms may operate in several sectors of the
these technologies accompanied the commercializa- economy, the firm is considered to be operating in
tion of the Internet and therefore would have been the sector that houses the largest share of its total
unlikely to have diffused widely prior to 1995. We employment. We find that 94% of all firms in the
utilize the information if a firm uses EDI in robust- merged LFTTD and CNUS data sets operate in the
ness checks as a control for the existence of previous manufacturing sector. This results in our analysis
network technology. The CNUS sample consists of sample of 5850 unique firms that engaged in inter-
about 38,000 plants belonging to about 22,000 national trade – related party, unrelated party, or
firms; together, they account for more than half of both – in 1992 and 1999.
manufacturing employment and output in the Uni-
ted States at the time (US Census Bureau, 1999). Measures

Intra-firm trade transactions data Dependent variable

We link the computer network data to a firm’s Since we are interested in the firm’s make or
international merchandise trade transactions in the buy decisions across international borders, we
LFTTD. We focus on intra-firm trade shares because operationalize the foreign boundary choice of MNEs
of the absence of detailed plant-level trade data in between in-house and arm’s length production by
the annual surveys or censuses of manufacturing. utilizing information on (1) intra-firm trade
The LFTTD is a transaction firm-linked database between US parents and their foreign affiliates as
linking the universe of individual merchandise trade well as between foreign-controlled US affiliates and
transactions, both exports and imports, to the US their foreign parent groups; and (2) inter-firm trade
firms that make them. We use information on the between US parents and other non-affiliated parties
value, date of transaction, and the categorical vari- as well as between US affiliates of foreign parents
able that indicates whether a trade transaction took and other non-affiliated parties.2 Specifically, the
place between related parties or not. Exporting dependent variable is the change in intra-firm trade
parties are related when either owns 10% or more of as a share of total trade value at the firm level
the other party, whereas for imports, the threshold between 1992 and 1999. We construct intra-firm
for related party is 5% under US Code 19, §1401a(g). trade shares using both import and export transac-
Ruhl (2013) discusses the comparability of the intra- tions in the LFTTD. The numerator is the sum of
firm trade measure derived from merchandise trade total exports and imports of the firm that are
transactions data compared with that derived from conducted between related parties, and the denomi-
Bureau of Economic Analysis (BEA) surveys and finds nator is the sum of total exports and imports
them mostly similar. The related party variable has that the firm conducts between both related and
previously been used to identify multinational activ- unrelated parties. Our proposed measure allows
ity by Bernard, Jensen, and Schott (2009) and Nunn us to contrast intra-firm and arm’s length trade
and Trefler (2013) among others. flows (Hipple, 1990). In particular, it is measured in
We first select firms in the LFTTD that are present levels as:
in both 1992 and 1999, since we adopt a difference-


in-differences methodology requiring us to follow IFit ¼ Exportsirt + Importsirt = Exportsit + Importsit ;
changes in intra-firm trade shares within a firm (1)

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
569

where i indexes firm, r related party, and t year. Forman (2005) considers basic Internet access as
The variable used in our empirical specification is simple while complex technologies include
therefore given by: business-to-customer and business-to-business
ΔIFi ¼ IFi;1999 - IFi;1992 (2) e-commerce. Following this logic, we rank the com-
puter networks reported in CNUS in ascending order
Note that our explicit assumption here is that the of complexity as Internet and all other Internet
firm has already made a decision whether or not to applications for e-commerce, namely, LAN, intranet,
engage in foreign production. We do not model the and extranet. Internetit is measured as in (3)
decision of a firm to become a multinational. where Networkjt takes on a value 1 if a plant chose
Internet as the type of network used as indicated
Independent variables in Question 1. We create similar measures
Our explanatory variable of interest is a firm-level as in (3) for LAN, intranet, and extranet sepa-
measure of ICT adoption, ICTit, bounded between rately. Then we create an indicator variable,
0 and 1. We first create an indicator variable, Ecommerce Applicationsit, that takes on a value of 1
Networkjt, that takes on a value 1 if plant j has any if a firm has any of the following – LAN, intranet, or
one of the target types of Internet-enabled computer extranet – and 0 otherwise. Finally, we have our
networks – the Internet, LAN, intranet, or extranet – measures of interest similar to (4) as ΔInterneti and
and 0 otherwise. This variable takes on a value ΔE-commerce Applicationsi, respectively.
of 0 for all plants in 1992 following the assumption Although Internet is the largest computer network
of little or no diffusion of Internet-enabled ICT pre- by definition, because of its global scope, it cannot
1995. Since none, few, or all of the plants belonging be customized by the firm. The other three types of
to a firm may adopt ICT, in order to create a firm- networks allow restricted access to various parties
level measure of ICT adoption, we weigh this indi- depending on the firm’s goals. LAN is a private
cator variable by the plant’s share in total firm network that links computers at a single location,
employment as follows: such as at an office building. Intranets and extranets
Xn   are both far more complex than LAN. Extranets
ICTi;1999 ¼ Employmentj;1999 =Employmenti;1999 allow controlled access – typically to customers or
j¼1 suppliers but also to geographically dispersed divi-
*Networkj;1999 ; ð3Þ sions of a company – from outside of an organiza-
tion’s intranet. We consider the Internet as the
where j indexes manufacturing plants belonging to simplest type of network since no particular control
firm i, and a firm may have n number of plants. For a or access infrastructure are required, while the rest
single-plant firm and multi-plant firm where all its are increasingly more complex in their configura-
plants adopt an Internet-enabled network technol- tions relative to the Internet.
ogy, the employment weight is one. Approximately, To assess the degree of electronic codifiability of
85% of the firms in our analysis sample are multi- the production process in industry k, Codifiabili-
plant firms, and in about half of these firms, some tyk, we calculate the fraction of plants in a six-digit
but not all of their plants adopt a network technol- North American Industrial Classification System
ogy. Therefore we create an employment weighted (NAICS) industry that used integrated CAD or
measure of network adoption at the firm level. We CAE business processes in 1999. Therefore this
also carry out robustness checks to ensure that our measure does not vary over time. We utilize Ques-
results are not being driven by this particular weight- tion 7, Part C, Number 1, in the CNUS to construct
ing scheme. In particular, we re-run our base specifi- the share of plants in an industry that uses CAD/
cation on a sample of single-plant firms only as well CAE relative to plants that did not use CAD/CAE
as multi-plant firms where all of the plants belong- and had no plans to use it by 2002, following Fort
ing to the firm adopt ICT. Finally, our dependent (2013). Both CAD and CAE are computer-aided
variable of interest, assuming all plants in 1992 had processes that may be used to translate manufac-
not adopted Internet-enabled ICT, is as follows: turing designs and specifications into electronic
ΔICTi ¼ ICTi;1999 : (4) formats. Therefore Codifiabilityk, measures the
intensity of CAD/CAE use in an industry and
In order to evaluate Hypothesis 1, we appeal to captures the relative degree to which industry
Forman (2005) in ranking the various types of production processes can be fragmented and codi-
networking technologies as simple or more complex. fied into electronic formats.

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
570

Table 1 Summary statistics (1992 and 1999)

Variable name Variable description Mean (standard deviation)

1992 1999

ICT 1 if a firm adopts an Internet-enabled computer network 0 otherwise NA 0.666 (0.323)

Types of ICT
Internet 1 if a firm adopts Internet 0 otherwise NA 0.574 (0.367)
E-commerce 1 if a firm adopts LAN, intranet, or extranet 0 otherwise NA 0.852 (0.355)
applications
Intra-firm Trade Share Intra-Firm (Exports + Imports)/Total Trade 0.133 (0.239) 0.143 (0.237)
Intra-firm Export Share Intra-Firm Exports/Total Trade 0.072 (0.158) 0.074 (0.149)
Intra-firm Import Share Intra-Firm Imports/Total Trade 0.061 (0.166) 0.070 (0.169)
Initial Age Firm Age as of 1992 12.568 (5.551) NA
Labor Productivity Total Value of Shipments/Total Employment 159.041 (214.421) 224.45 (236.553)
Observations 5850 5850
Notes: Internet-enabled computer networks are defined as Internet, LAN, extranet, or intranet from Question 1 in the CNUS. Total trade refers to both
exports and imports. Number of observations rounded to the nearest integer for disclosure avoidance.

Control variables increases from $159,000 per employee to $224,000

We additionally include firm specific variables of age per employee between 1992 and 1999.
as of 1992, Initial Agei, and change in labor produc-
tivity between 1999 and 1992, ΔLaborProductivityi, ECONOMETRIC MODEL
to control for experience and productivity of the We use a fixed effects panel data model as in Forman
firm. We obtain information on firm age from the and McElheran (2012), thus implementing a
LBD. For firms with multiple plants, age is calculated difference-in-differences identification strategy. We
as the difference between the year of interest and the have a simple set-up with two time periods, 1992
year of establishment of its oldest plant. We consider and 1999, and two distinct groups of firms, those
age as of 1992, since in our difference-in-differences that adopt ICT (treated firms) and those that do not
specification, the change in age will be a constant (control firms). We compare the share of intra-firm
7 years for all firms. Labor productivity is measured trade in overall trade prior to adoption of ICT to the
as total value of shipments per employee and we share after adoption, and relative to those that
measure the change between 1992 and 1999. Firm- do not adopt ICT. Thus the first difference is the
level measures of shipments and employment are difference in intra-firm trade shares between 1992
constructed as the sum of these respective variables and 1999 and the second being the difference
over all plants belonging to the firm. The plant-level between ICT adopters and non-adopters. Our key
data on the value of shipments and employment in identification assumption is that diffusion of
1992 and 1999 is sourced from the 1992 Census of Internet-enabled ICT was not widespread among
Manufactures and the 1999 ASM, respectively. firms prior to 1995 (Forman et al., 2005). Therefore
Table 1 provides summary statistics of all our main we choose the earliest year for which trade data is
variables in 1992 and 1999. We can see that about available as the pre-period, 1992, at which time all
67% of the firms in our sample adopt ICT. This share firms, following previous studies (Forman et al.,
varies by the various types of ICT. About 57% of the 2005; Forman & van Zeebroeck, 2012), are assumed
firms adopt the Internet, while a larger share, 85%, to have no or narrowly diffused Internet-enabled
adopt the LAN, intranet, and/or extranet. The table ICT technology.
also provides the average intra-firm trade shares Our empirical approach removes firm-level time-
further differentiated by exports and imports. We invariant characteristics that may be correlated
see that in 1992, 13% of total trade is intra-firm with both ICT adoption and intra-firm trade shares.
in nature for an average firm in our sample with However, time-varying firm characteristics such as
about an even split between intra-firm exports and productivity may influence the decisions to simul-
imports. This share rises to over 14% by 1999. taneously engage in intra-firm trade and adopt ICT.
Finally, average age as of 1992 of a firm in our sample We include controls for firm productivity and age
is about 13 years and average labor productivity to address such concerns about omitted variables.

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
571

Table 2 Correlation matrix (1992 and 1999)

Variable (1) (2) (3) (4) (5) (6)

1. Change in Intra-firm Trade Share 1.00 — — — — —

2. Change in ICT 0.035 1.00 — — — —
3. Change in Internet 0.029 0.868 1.00 — — —
4. Change in E-commerce applications 0.038 0.739 0.619 1.00 — —
5. Initial Age −0.022a 0.016b 0.010c −0.018a 1.00 —
6. Change in Labor Productivity 0.049 0.088 0.085 0.136 −0.019a 1.00
Observations 11,700
a
Significant at 5% level.
b
Significant at 10% level.
c
Not significant. Number of observations rounded to the nearest integer for disclosure avoidance.
Notes: All pairwise correlations are significant at the 1% level, unless otherwise noted.

Since we estimate a two period difference-in-differ- run the following specification:

ences model, we implement the following regression ΔIFi ¼β1 ΔICTi + β2 ΔICTi *Codifiabilityk
equation:
+ β3 ΔLabor Productivityi
ΔIFi ¼β1 ΔICTi + β2 ΔLabor Productivityi + β4 Initial Agei;1992 + Δεi ; ð7Þ
+ β3 Initial Agei;1992 + Δεi ; ð5Þ
where the industry codifiability measure is con-
structed based on the fraction of plants in an indus-
where Δ denotes change between t and t−1 such that
try, k, that used CAD/CAE software in their
ΔIFi = IFi,1999 − IFi,1992 represents change in intra-
production process in 1999. Our interest is in the
firm trade shares between 1999 and 1992 and simi-
interaction term and we expect β2 to be negative.
larly for the explanatory variables. We rely on the
data to reveal the sign of β1. RESULTS
For testing Hypothesis 1, we separately include
specific measures of ICT, that is, ΔInterneti and Baseline Results
ΔE-commerce Applicationsi as follows: Table 2 presents a correlation matrix of the main
variables over our sample period. Since we employ
ΔIFi ¼β1 ΔInterneti + β2 ΔLabor Productivityi changes over time in our empirical specifications, we
+ β3 Initial Agei;1992 + Δεi ; ð6aÞ provide correlations between the change measure of
a variable rather than in levels, except for initial firm
age that is calculated as of 1992. We find that ICT
ΔIFi ¼β1 ΔEcommercei + β2 ΔLabor Productivityi and each of its types separately are positively and
+ β3 Initial Agei;1992 + Δεi ; ð6bÞ statistically significantly correlated with changes in
intra-firm trade shares. The overall ICT adoption
Abstracting from the initial decision of whether or measure is highly correlated with the adoption of
not to adopt ICT, the ability to maximize the the Internet and other e-commerce applications.
benefits of having complex ICT systems would be This is not surprising given that our overall ICT
easier between two parties that have more integra- measure is comprised of the Internet, LAN, intranet,
tion and coordination, and thus favor in-house and extranet.
production, which should be reflected in increased In Table 3 we present regression results to examine
intra-firm transactions relative to arm’s length. Thus the net effect of ICT on MNE foreign boundary
we hypothesize that the sign of β1 will be larger in choices. The estimates reported in the tables reflect
Eq. (6b) compared with Eq. (6a). the average difference in intra-firm trade shares
To assess Hypothesis 2, the extent to which the between ICT adopters and non-adopters, net the
impact of ICT on intra-firm trade shares varies in initial difference in the pre-period. The first column
industries with varying degrees of electronic codifia- of Table 3 displays results from specification (5). We
bility, that is, propensity to codify product specifica- see that overall ICT adoption at the firm level is
tions in an electronic format, and thus easily associated with a 1.8 percentage point increase in
communicate with parties in foreign locations, we intra-firm trade shares compared with those MNEs

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
572

Table 3 Difference-in-Differences regressions explaining intra-firm trade shares: The role of ICT

Variable (1) (2) (3)

ICT 0.018** (0.006)

Types of ICT
Internet 0.026** (0.005)
E-commerce applications 0.032*** (0.005)
Initial Age −1.030* (0.000) −0.022* (0.000) −0.021* (0.000)
Labor Productivity 0.033** (0.000) 0.046** (0.000) 0.044** (0.000)
Observations 11,700 11,700 11,700
Notes: Significance level: * if p<0.05; ** if p<0.01; *** if p<0.001. Robust standard errors in parentheses. All regressions are estimated in first differences;
initial age included in levels. Columns (2) and (3) report β coefficients. Age and Labor Productivity coefficients multiplied by 1000 for ease of interpretation.
Number of observations rounded to the nearest integer for disclosure avoidance.

that do not adopt ICT, net of the difference in the Table 4 Difference-in-Differences regressions explaining intra-
initial period. Our result suggests that, for MNEs firm trade shares: The role of ICT and industry electronic
operating in the US manufacturing sector, ICT- codifiability
induced internal cost reductions outweigh external Variable (1)
cost reductions, thus favoring internal cross-border
transactions over external ones. ICT 0.021*** (0.007)
To test Hypothesis 1, we consider various types of ICT × Codifiability −0.002† (0.001)
Initial Age −0.066 (0.000)
ICT separately in Columns 2 and 3 following speci-
Labor Productivity 0.033** (0.000)
fications (6a) and (6b), respectively. We find that
Observations 11,640
more complex Internet-enabled ICT tends to have a
larger impact on internal cross-border transactions, Notes: Significance level: † if p<0.10, * if p<0.05; ** if p<0.01; *** if
p<0.001. Robust standard errors in parentheses. All regressions are
as exhibited by increased intra-firm trade shares. estimated in first differences; initial age included in levels. “Codifiability”
In order to compare the magnitudes of the coeffi- is measured as the share of plants that use CAD/CAE technology in a six-
digit NAICS industry. Age and Labor Productivity coefficients multiplied by
cients on our Internet and e-commerce application 1000 for ease of interpretation. Number of observations rounded to the
variables, we compute and report β coefficients. nearest integer for disclosure avoidance.
β coefficients refer to how many standard deviations
a dependent variable will change, per standard
deviation increase in the explanatory variable, Internet-enabled ICT adoption on intra-firm trade.
thereby allowing us to compare the magnitude of Although ICT adoption itself is associated with
coefficients across different regression specifications. higher intra-firm trade shares, in industries more
We find that the β coefficient on the Internet prone to electronic codifiability of its production
variable is 0.026 while the e-commerce applications process, this effect is muted.
variable is 0.032. This implies that a one standard In all our specifications, the coefficients on the
deviation change in Internet use is associated with firm-level control variables of age and labor produc-
an increase of 0.026 in intra-firm trade shares, while tivity are similar. Initial firm age is negative and
the increase is 0.032 when considering e-commerce statistically significant implying that older firms in
applications. These findings together provide sup- our sample tend to experience decreased intra-firm
port for Hypothesis 1, in that more complicated trade over the sample period. The coefficient on
forms of ICT adoption that require simultaneous labor productivity is positive and statistically signifi-
investments in ICT infrastructure encourage more cant implying that changes in firm productivity are
internal firm activities, highlighted by increased positively correlated with changes in intra-firm trade
trade between internal firm units relative to arm’s shares.
length trade.
Next, Table 4 presents the results evaluating Robustness Checks
Hypothesis 2 following specification (7). The num- We conduct several robustness checks to further test
ber of firms in the analysis sample is reduced because for simultaneity bias and to check whether our
we were not able to create a measure of electronic results are sensitive to sample selection. In order to
codifiability for every six-digit NAICS represented in address concerns that firms may proactively adopt
our sample. As predicted, the degree of electronic ICT in anticipation of enhancing intra-firm or arm’s
codifiability in an industry dampens the impact of length trade, we re-run our baseline specification

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
573

using information on a firm’s use of EDI. EDI is an trade patterns. We find that the adoption of ICT to
older and proprietary network technology used to carry out each of these functions has a positive effect
exchange documents between businesses in a stan- on intra-firm trade shares.
dardized format. It is one of the choices of computer Since we can distinguish between related party
networks included in Question 1 of the CNUS. The exports and related party imports in our data, we
existence of EDI indicates prior investments in further explore the role of imports or exports in
information and communication technology at a driving the overall results. To this end, we rerun the
firm. As Forman (2005) finds, one of the important regressions using only intra-firm exports as a share of
factors determining likelihood of Internet adoption total trade, and then separately using only intra-firm
is prior network investment. We re-run our baseline imports as a share of total trade. The results again are
results controlling for the existence of EDI that is statistically significant and similar to those using
constructed as outlined in Eq. (3) and find that total intra-firm trade shares, suggesting that ICT
results still remain quantitatively and qualitatively adoption impacted both intra-firm exports and
unchanged. imports similarly. All robustness check results are
Due to the aggregation of plant-level data to the available upon request.
firm level, the results may be influenced by the We acknowledge that even after conducting the
weighting mechanism that we employ. To account above-discussed robustness checks, the difference-
for that, we re-run our econometric specifications on in-differences methodology implemented in this
a sample consisting of only firms that are single article may be subject to endogeneity concerns. In
plants in the United States. The reasoning here is particular, aside from controlling for prior technol-
that we are not using any weights in constructing ogy investment at a firm, we are not fully able to
the ICT measures, since the firm and plant is synon- control for all time-varying factors that may simulta-
ymous. The results are virtually identical to those neously impact ICT adoption and intra-firm trade
using the whole sample including both multi- and shares. However, in the absence of a longer panel of
single-plant firms. Next, in cases where only a few the CNUS or other similar surveys, we utilize this
plants within multi-plant firms adopt ICT, the iden- unique data to provide valuable insights into the
tification of adoption vs non-adoption at the firm impact of ICT adoption on foreign boundary choices
level is not strict. Thus we impose a stricter defini- for a group of early technology adopters.
tion of adoption and non-adoption, similar to con-
sidering single-unit firms only, but including only DISCUSSION AND CONCLUSIONS
multi-plant firms where all or none of its plants We analyzed the impact of ICT adoption on MNEs’
adopt ICT. In this subsample, the explanatory vari- foreign boundary choices. Through the lens of trans-
able is a purely categorical variable that takes on a actions cost theory, we argue that the choices of
value 1 when all plants within a firm adopt ICT and arm’s length and intra-firm trade within the MNE
0 if none of the plants within a firm adopt ICT. The are not mutually exclusive, but rather complemen-
results are robust to these sample variations. tary. ICT can lower search and communications
It could also be the case that ICT has a stronger costs between the focal firm and both its internal
influence on certain functions within the firm. The and external trading partners. However, we high-
CNUS includes various questions about the stage of light that as the complexity of the adopted ICT
the production process that utilizes ICT. We follow increases, it requires matching capabilities and
the survey’s breakdown in areas of plant operation mutual commitment in the trading partner’s ICT
(Question 7): purchasing, orders of manufactured adoption, hence increasing the likelihood of intra-
products, production management, logistics, and firm trade. At the same time, ICT adoption tends
communication and support services. Under each of to favor external transactions in industries where
these categories, the adoption of ICT is positive and the production processes are readily translated into
statistically significant with respect to its relation- electronic formats. For firms operating in these
ship with intra-firm trade shares. The results support industries, ICT adoption facilitates the fragmenta-
Keane and Feinberg’s (2007) finding that improve- tion of production processes. This breaking up of the
ments in logistics encouraged intra-firm trade production process into smaller value chains allows
between US parents and their Canadian affiliates. the firm to control the extent of intellectual property
Yet more importantly, it also highlights that logistics shared at each stage of the production and to partner
is not the only area of production that was affected with the least cost provider at each juncture of
by ICT adoption and subsequently a firm’s intra-firm the supply chain, thus creating more channels of

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
574

contracting with external partners. Regression following ICT adoption. Our study is silent on the
results confirm our hypotheses. role of ICT adoption on the extensive margin decision
Our findings contribute to two main strands of – should a firm choose to become or cease to exist as a
literature, that is, one on information technology multinational. We are not alone, however, in making
and firm boundaries, and a second one on offshoring this assumption. Feinberg and Keane (2006) employ
and global value chains. While the first strand data from the Benchmark and Annual Surveys of US
has been intensively studied in the domestic con- Direct Investment Abroad administered by the BEA,
text (Bloom et al., 2014; Brynjolfsson, Malone, which contains the population of US-based MNEs,
Gurbaxani, &, Kambil, 1994; Forman & McElheran, thus, by virtue of being included in this data set, a
2012), the link to the international dimension of firm must be an MNE. Rangan and Sengul (2009)
firm boundary has received less attention (excep- make use of industry-level data set of US manufactur-
tions are Feinberg & Keane (2006), Keane & Feinberg ing MNEs in their study of ICT and transnational
(2007), Rangan & Sengul (2009)). In the burgeoning integration. Future studies could explore the timing of
literature on offshoring and global value chains both ICT adoption and becoming an MNE.
(Gereffi, Humphrey, & Sturgeon, 2005), most exist- Moreover, our sample period represents a time in
ing studies implicitly and explicitly make the which US manufacturing sectors adopted major,
assumption that ICT developments and adoption yet basic, Internet-enabled ICT advancements that
have contributed to the process of the slicing and changed production processes in fundamental ways.
dicing of the production chain. In this article we Future research examining current developments in
establish a specific link between Internet-enabled ICT and the effect of more sophisticated technolo-
ICT adoption and intra-firm trade shares. Our within gies on firm’s foreign boundary choices could build
firm-level measures of detailed ICT adoption and on the insights from our results. Recent develop-
trade patterns allows analysis at a highly disaggre- ments in digital manufacturing – such as additive
gated level that can account for the heterogeneity in manufacturing, continuous manufacturing, collec-
ICT adoption. Moreover, our relatively early period tive manufacturing, crowdsourcing, cloud comput-
of study coincides with the first significant wave of ing and cloud manufacturing, and 3D printing –
Internet-enabled ICT adoption (Forman et al., 2005; have revolutionized production processes and cre-
Forman & van Zeebroeck, 2012) that is still the basis ated new outlets for global value chain production.
for many types of ICT that are being introduced In particular, future studies could examine how
today. Since these types of Internet-enabled technol- specific types of ICT adoption may lead to upgrad-
ogy had just become available around that time ing of the global value chain and affect global
period, our analysis provides a valuable benchmark value chain governance (see Gereffi et al., 2005 for
that can act as a baseline and lower bound of the a comprehensive overview of global value chain).
effects of ICT adoption on a firm’s global production As ICT itself is subject to constant changes and devel-
patterns. Our results also show that the propensity opment, it would be worthwhile to assess the effects of
for fragmented production in industries favors more recent developments in ICT on MNEs’ boundary
outsourcing by firms following ICT adoption. The choices, which we leave for future research.
advent of CAD and manufacturing has been pro-
posed as an explanation for the relative intensity
with which activities stay within the confines of the ACKNOWLEDGEMENTS
firm boundary and the subsequent nature of trade The authors would like to thank the Special Issue Editors
patterns (Helpman, 2006). Our results highlight that Juan Alcacer, John Cantwell, and Lucia Piscitello, as
ICT adoption acts as a catalyst to exploit the extent well as four anonymous reviewers, for their helpful
to which production processes may be translated comments. They are especially indebted to comments
into electronic formats leading to more complex trade by Rene Belderbos, Alvaro Cuervo-Cazurra, Jose de la
patterns, with the caveat that we focus on a sample of Torre, Susan Feinberg, Cheryl Grim, Anupama Phene,
firms operating in the US manufacturing sector. and Nikolas Zolas that have greatly helped in shaping
Having highlighted the contributions of our study, this article. They would also like to thank participants
we acknowledge that more remains to be done. at the JIBS Special Issue conference and at the 2014
The explicit assumption that firms have already Academy of International Business Annual Meeting for
decided on foreign production limits our results in useful comments and suggestions. Chen thanks the
that we focus solely on the intensive margin Institute for International Economic Policy at GWU for
decision – how much does intra-firm trade change financial support. All errors remain of the authors.

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
575

NOTES 70% (60%) of intra-firm exports (imports) are

1 between US parents (foreign parents) and their
References to specific companies in the text are
based on public information. foreign affiliates (US affiliates) in 1992 and 1999
2 (Zeile, 2003).
Although we cannot distinguish between US and
foreign MNEs in the trade transactions data, about

REFERENCES
Antràs, P., & Yeaple, S. R. 2014. Multinational firms and the Feinberg, S. E., & Keane, M. P. 2006. Accounting for the growth
structure of international trade. In E. Helpman, K. Rogoff, & G. of MNC-based trade using a structural model of US MNCs.
Gopinath (Eds), Handbook of international economics: 55–130. American Economic Review, 96(5): 1515–1558.
Amsterdam: Elsevier. Forman, C. 2005. The corporate digital divide: Determinants of
Antràs, P., Rossi-Hansberg, E., & Garicano, L. 2008. Organizing internet adoption. Management Science, 51(4): 641–654.
offshoring: Middle managers and communication costs. In Forman, C., Goldfarb, A., & Greenstein, S. 2005. How did
E. Helpman, T. Verdier, & D. Marin (Eds), The organization of location affect adoption of the commercial internet? Global
firms in a global economy: 311–339. Cambridge, MA: Harvard village vs. urban leadership. Journal of Urban Economics, 58(3):
University Press. 389–420.
Baldwin, R., & Lopez-Gonzalez, J. 2014. Supply-chain trade: Forman, C., & McElheran, K. 2012. Information technology and
A portrait of global patterns and several testable hypotheses. boundary of the firm: Evidence from plant-level data. Harvard
World Economy, 38(11): 1682–1721. Business School Working Paper 12-092.
Bernard, A. B., Jensen, J. B., & Schott, P. K. 2009. A portrait of Forman, C., & van Zeebroeck, N. 2012. From wires to partners:
firms in the US that trade goods. In T. Dunne, J. B. Jensen, & How has the internet fostered R&D collaborations within firms.
M. J. Roberts (Eds), Producer dynamics: New evidence from Management Science, 58(8): 1549–1568.
micro data: 383–410. Chicago, IL: University of Chicago Fort, T. C. 2013. Breaking up is hard to do: Why firms fragment
Press. production across locations. Working Papers 13-35, Center for
Bloom, N., Garicano, L., Sadun, R., & Van Reenen, J. 2014. The Economic Studies, US Census Bureau.
distinct effects of information technology and communication Gereffi, G., Humphrey, J., & Sturgeon, T. 2005. The governance
technology on firm organization. Management Science, 60(12): of global value chains. Review of International Political Economy,
2859–2885. 12(1): 78–104.
Buckley, P. J., & Casson, M. C. 1976. The future of the multinational Grossman, G., & Rossi-Hansberg, E. 2008. Trading tasks:
enterprise. London: Palgrave Macmillan. A simple theory of offshoring. American Economic Review,
Buckley, P. J., & Casson, M. C. 1998. Analyzing foreign market 98(5): 1978–97.
entry strategies: Extending the internalization approach. Gurbaxani, V., & Whang, S. 1991. The impact of information
Journal of International Business Studies, 29(3): 539–561. systems on organizations and markets. Communications of the
Buckley, P. J., & Casson, M. C. 2009. The internalisation theory of ACM, 34(1): 59–73.
the multinational enterprise: A review of the progress of a Helpman, E. 2006. Trade, FDI, and the organization of firms.
research agenda after 30 years. Journal of International Business Journal of Economic Literature, 44(3): 589–630.
Studies, 40(9): 1563–1580. Hennart, J.-F. 1982. A theory of multinational enterprise. Ann Arbor,
Brynjolfsson, E. 1994. Information assets, technology, and MI: University of Michigan Press.
organization. Management Science, 40(12): 1645–1662. Hennart, J.-F. 2009. Down with MNE-centric theories! Market
Brynjolfsson, E., Malone, T., Gurbaxani, V., & Kambil, A. 1994. entry and expansion as the bundling of MNE and local assets.
Does information technology lead to smaller firms? Manage- Journal of International Business Studies, 40(9): 1432–1454.
ment Science, 40(12): 1645–1662. Hipple, F. S. 1990. The measurement of international trade
Cho, K. R. 1990. The role of product-specific factors in intra-firm related to multinational companies. American Economic Review,
trade of US manufacturing multinational corporations. Journal 80(5): 1263–1270.
of International Business Studies, 21(2): 319–330. Hitt, L. 1999. Information technology and firm boundaries:
Clemens, E. K., Reddi, S. P., & Row, M. C. 1993. The impact Evidence from panel data. Information Systems Research, 10(2):
of information technology on the organization of economic 134–149.
activity: The ‘move to the middle’ hypothesis. Journal of Jarmin, R. S., & Miranda, J. 2002. The longitudinal business
Management Information Systems, 10(2): 9–35. database. Working Papers 02-17, Center for Economic Studies,
Costinot, A., Vogel, J., & Wang, S. 2013. An elementary theory US Census Bureau.
of global supply chains. Review of Economic Studies, 80(1): Jones, R., & Kierzkowski, H. 2001. A framework for fragmenta-
109–144. tion. In S. Arndt, & H. Kierzkowski (Eds), Fragmentation:
De la Torre, J., & Moxon, R. W. 2001. Introduction to the New production patterns in the world economy. Oxford: Oxford
symposium e-commerce and global business: The impact of University Press.
the information and communication technology revolution on Keane, M. P., & Feinberg, S. E. 2007. Advances in logistics and
the conduct of international business. Journal of International the growth of intra-firm trade: The case of Canadian affiliates of
Business Studies, 32(4): 617–639. US multinationals, 1984–1995. Journal of Industrial Economics,
Deardorff, A. V. 2001. Fragmentation across cones. In S. Arndt, & 55(4): 571–632.
H. Kierzkowski (Eds), Fragmentation: New production patterns in Kim, S. M., & Mahoney, J. T. 2005. Property rights theory,
the world economy 35–51. New York: Oxford University Press. transaction costs theory, and agency theory: An organizational
Dosi, G., Gambardella, A., Grazzi, M., & Orsenigo, L. 2008. economics approach to strategic management. Strategic
Technological revolutions and the evolution of industrial Management Journal, 25(5): 489–496.
structures: Assessing the impact of new technologies upon Kohler, W. 2004. International outsourcing and factor prices
the size and boundaries of firms. Capitalism and Society, 3(1): with multistage production. The Economic Journal, 114(494):
1–49. C166–C185.
Dunning, J. H. 1980. Toward an eclectic theory of international Malone, T. W., Yates, J., & Benjamin, R. I. 1987. Electronic markets
production: Some empirical tests. Journal of International Busi- and electronic hierarchies. Communications of the ACM, 30(6):
ness Studies, 11(1): 9–31. 484–497.

Journal of International Business Studies

 The impact of ICT adoption Wenjie Chen and Fariha Kamal
576

McPhee, R., & Poole, M. S. 2001. Organizational structures US Census Bureau. 1999. US manufacturers’ value of shipments,
and configurations. In F. M. Jablin, & L. L. Putnam (Eds), The www.census.gov//econ/estats/1999/manu_initial/MCDtables.
new handbook of organizational communication: Advances in pdf, accessed 13 August 2014.
theory, research, and methods: 503–543. Thousand Oaks, CA: Worthen, B. 2002. Nestlé’s ERP Odyssey. CIO Magazine, 62–70.
Sage. Zeile, W. J. 2003. Trade in goods with multinational companies:
Messenbourg, T. 2001. Measuring electronic business definitions, Survey-based data and findings for the United States of America,
underlying concepts, and measurement plans. US Census Bureau, www.bea.gov/papers/pdf/IFT_OECD_Zeile.pdf, accessed 13
www.census.gov/econ/estats/papers/ebusasa.pdf, accessed August 2014.
13 August 2014. Zenger, T. R., & Hesterly, W. S. 1997. The disaggregation of
McElheran, K. 2015. Do market leaders lead in business process corporations: Selective intervention, high-powered incentives,
innovation? The case(s) of e-business adoption. Management and molecular units. Organization Science, 8(3): 209–222.
Science, 61(6): 1197–1216.
Nunn, T., & Trefler, D. 2013. Incomplete contracts and bound-
aries of the multinational firm. Journal of Economic Behavior and
Organization, 94(1): 330–344. ABOUT THE AUTHORS
Rangan, S., & Sengul, M. 2009. Information technology and Wenjie Chen is an Assistant Professor of Inter-
transnational integration: Theory and evidence on the evolu-
tion of modern multinational enterprise. Journal of International national Business at the School of Business at George
Business Studies, 40(9): 1496–1514. Washington University. She holds a PhD in Eco-
Rice, R. E., & Gattiker, U. E. 2001. New media and organizational
structuring. In F. M. Jablin, & L. L. Putnam (Eds), The new
nomics from the University of Michigan. Her area of
handbook of organizational communication: Advances in theory, research is in foreign direct investment, cross-border
research, and methods: 544–581. Thousand Oaks, CA: Sage. mergers and acquisitions, trade, and applied econo-
Ruhl, K. J. 2013. An overview of US intrafirm-trade data sources.
Mimeo. metric methods.
Sturgeon, T. J. 2002. Modular production networks: A new
American model of industrial organization. Industrial and Cor- Fariha Kamal is a Senior Research Economist at
porate Change, 11(3): 451–496.
Teece, D. J. 1986. Transactions cost economics and the multi- the Center for Economic Studies at the US Census
national enterprise. Journal of Economic Behavior and Organiza- Bureau. She has a MBA from Clark University and
tion, 7(1): 21–45. PhD in Economics from Syracuse University. Her
UNCTAD. 2013. World investment report, http://unctad.org/
en/pages/PressRelease.aspx?OriginalVersionID=113, accessed research interests are in the areas of international
13 August 2014. trade, development, and economic geography.

Disclaimer: Any opinions and conclusions expressed herein are those of the author(s) and do not necessarily
represent the views of the US Census Bureau. All results have been reviewed to ensure that no confidential
information is disclosed.

Accepted by Juan Alcacer, Guest Editor, 29 December 2015. This article has been with the authors for four revisions.

Journal of International Business Studies