Globalization and Offshoring of Software Full Final

Download PDF


Association for Computing Machinery

Advancing Computing as a Science & Profession

Globalization and

Offshoring of Software

A Report of the ACM Job Migration Task Force

For more information about ACM,


For an electronic version of all or parts of this report,


The Association for Computing Machinery

(ACM) is an educational and scientific society

uniting the world’s computing educators,

researchers, and professionals to inspire

dialogue, share resources, and address the

field’s challenges. ACM strengthens the

profession’s collective voice through strong

leadership, promotion of the highest standards,

and recognition of technical excellence. ACM

supports the professional growth of its

members by providing opportunities for lifelong

learning, career development, and

professional networking.

Globalization and

Offshoring of Software

A Report of the ACM Job Migration Task Force

William Aspray, Frank Mayadas, and Moshe Y. Vardi, Editors

Association for Computing Machinery

Advancing Computing as a Science & Profession

Copyright © 2006 ACM 0001-0782/06/0200. Permission to make digital or hard copies of all or part of this work for

personal or classroom use is granted without fee provided that copies are not made or distributed for profit or

commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to

republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

Page 2

The ACM Job Migration Task Force

Alok Aggrawal (Founder and Chairman of Evalueserve)

Frances E. Allen (Fellow Emerita at the T. J. Watson Research Center)

Stephen J. Andriole (Thomas G. LaBrecque Professor of Business Technology at Villanova


Ashish Arora (Professor of Economics and Public Policy at Carnegie Mellon University)

William Aspray (Rudy Professor of Informatics and Special Advisor on Information

Technology and Professional Partnerships in the Office of the Vice President for Research at

Indiana University)

G. Balatchandirane (Teacher of East Asian Economic History and Japanese language at

the Department of East Asian Studies, University of Delhi)

Burt S. Barnow (Associate Director for research and Principal Research Scientist at the

Institute for Policy Studies at the Johns Hopkins University)

Orna Berry (Venture Partner in Gemini Israel Funds Ltd. and the Chairperson of Adamind

(LSE: ADA) Ltd. and Prime Sense, Inc.)

Michael Blasgen (Consultant in information technology)

Arndt Bode (Professor of Informatics of Technische Universität München, Germany)

Jean Camp (Associate Professor in the School of Informatics, Adjunct Professor of

Telecommunications, and Adjunct Professor of Computer Science at Indiana University)

Seymour E. Goodman (Professor of International Affairs and Computing at the Sam Nunn

School of International Affairs and the College of Computing, Georgia Institute of


Vijay Gurbaxani (Associate Dean, Professor of Information Systems, and Director of the

Center for Research on IT and Organizations Industry-University Consortium at the

Graduate School of Management, University of California at Irvine)

Juris Hartmanis (Walter R. Read Professor of Computer Science and Engineering, Emeritus

at Cornell University)

Charles House (Director of Societal Impact of Technology for Intel Corporation)

William Jack (Corporate Vice President with SAIC and currently acting as the senior

telecommunications advisor to the Chief of IT IS at the National Security Agency)

Martin Kenney (Professor in the Department of Human and Community Development at

the University of California, Davis and a Senior Project Director at the Berkeley Roundtable

on the International Economy)

Stefanie Ann Lenway (Dean of the College of Business Administration at the University of

Illinois at Chicago, as well as Professor of Management)

Vivek Mansingh (A Country Manager for India Development Center of Dell Products Group,


Page 3

Göran Marklund (Science and Technology Attaché at the Swedish Offices of Science and

Technology within the Swedish Embassy in Washington DC)

Frank Mayadas (Program Officer at the Alfred P. Sloan Foundation)

Peter Mertens (Professor for information systems at the University of Erlangen-

Nuremberg, Germany)

Rob Ramer (Information security consultant)

Bobby Schnabel (Vice Provost for Academic and Campus Technology at the University of

Colorado at Boulder)

Bankim Shah (Founder and president of BRS Associates, Inc)

Marie Stella (Lead security engineer for the FAA's National Airspace Communications


Valerie E. Taylor (Head of the Dwight Look College of Engineering's Department of

Computer Science and holder of the Royce E. Wisenbaker Professorship II at Texas A & M


Takashi Umezawa (Professor of Human Resource Management in Kokushikan University,

Tokyo, Japan)

Moshe Y. Vardi (George Professor in Computational Engineering and Director of the

Computer and Information Technology Institute at Rice University)

Roli Varma (Regents Lecturer and an associate professor in the School of Public

Administration at the University of New Mexico)

Richard C. Waters (President and CEO of Mitsubishi Electric Research Laboratories


John White (Executive Director and Chief Executive Officer of the Association for

Computing Machinery (ACM))

Stuart Zweben (Associate Dean of Academic Affairs and Administration in the College of

Engineering at The Ohio State University)

Page 4

Globalization and Offshoring of Software

A Report of the ACM Job Migration Task Force


For the past six decades, the Association for Computing Machinery (ACM) has been an

integral part of the evolution of computing as a science and profession. In early 2004, ACM

members began expressing concern about the future of computing as a viable field of study

and work. There were daily stories in national and international media describing major

shifts in employment that were occurring largely as a result of offshoring. Combined with

the impact of the end of the boom, these reports raised more questions than they

answered in the minds of many ACM members.

Given these concerns, ACM Council commissioned a Task Force to look at the facts behind

the rapid globalization of IT and the migration of jobs resulting from outsourcing and

offshoring. Being an international organization, ACM expected the task force to look at the

issue from a global perspective, as compared to a country-centric one. This was not

intended to be a study of offshoring from the United States to India and China and the

impact of that offshoring on the computing profession in the United States. Instead, the

task force was charged with looking at the forces shaping the migration of jobs worldwide in

the computing and information technology fields. Prior to this effort, no study has looked at

offshoring on a global scale.

ACM Presidents Maria Klawe (2002-04) and David Patterson (2004-06) invited Frank

Mayadas of the Sloan Foundation, Moshe Y. Vardi of Rice University, and Bill Aspray of

Indiana University to lead the effort. This group commissioned a task force of computer

scientists, social scientists, and labor economists from around the world. The Task Force

held four in-person meetings at which the facts and data surrounding the issue were

presented and discussed. In the process, trends emerged, myths were debunked, and a

more realistic picture of the current state and likely future of the information technology

field, profession, and industry emerged.

The report resulting from this study is significant. Moreover, the annotated bibliography

available on the ACM Web site provides the most comprehensive list of reports, resources,

and papers assembled on the topic of offshoring. As described in detail in the eight chapters

that comprise the report, the field of computing and information technology has experienced

a dramatic shift in the past five years to a truly global industry. The forces that have driven

and shaped this change are still at play and will continue. The implications for every ACM

member are significant. Full participation in the systems, software, and services portion of

the global information technology field will require deep grounding in the fundamentals of

computing, new knowledge surrounding business processes and platforms, and a deeper

understanding of the global community in which work will be done. The educational systems

that underpin our profession will need to change.

The future of IT is exciting, but it is a future very different from the past, and even from

the present.

John R. White

ACM Chief Executive Officer

Page 5


EXECUTIVE SUMMARY AND FINDINGS........................................................... 8

INTRODUCTION ............................................................................................14

OVERVIEW ....................................................................................................19

CHAPTER 1. OFFSHORING: THE BIG PICTURE...............................................44

What Do We Mean by Outsourcing, Offshoring, and Globalization

of Software? .................................................................................................45

How Did Offshoring Come About? ......................................................................46

How Much Work Is Offshored? ...........................................................................51

Which Countries Send Work Offshore and Which Countries Do

Most of This Work?........................................................................................52

What Types of Work Are Sent Offshore? .............................................................54

Why Are Firms Interested In Sending Work Offshore? ..........................................56

What Are the Technical, Business, and Other Drivers and Enablers

of Offshoring? ...............................................................................................60

Is IT Still a Good Career Choice for People Working in Countries That Ship IT

Jobs Overseas? .............................................................................................65

CHAPTER 2. THE ECONOMICS OF OFFSHORING ............................................71

The Economics of Offshoring: Rationale and Potential Impacts ..............................71

Data on the Current State of Offshoring and Projections for the Future ..................76

Conclusions .....................................................................................................94

CHAPTER 3. THE COUNTRY PERSPECTIVE ..................................................101

The History of Software Offshoring...................................................................102

The United States...........................................................................................104




The European Union—Western and Eastern Europe............................................124

Page 6


Conclusions ...................................................................................................129


Offshoring Firms ............................................................................................136

Overall Conclusion..........................................................................................152

CHAPTER 5. THE GLOBALIZATION OF IT RESEARCH ...................................155

Worldwide Distribution of IT Research ..............................................................159

Researcher Migration......................................................................................170

Research Job Migration ...................................................................................173


CHAPTER 6. OFFSHORING, RISKS AND EXPOSURES ...................................182

Vulnerabilities: Data and Network Security and Beyond......................................183

Corporate Risks and Information Security .........................................................186

Risks to the Individual: Privacy and Identity Theft .............................................191

Risks for National Capabilities and National Sovereignty .....................................201

Risk Mitigation and Risk Assessment ................................................................207

CHAPTER 7. EDUCATION .............................................................................213

Prospects and Challenges of an Educational Response to Offshoring ....................213

Indian Education ............................................................................................216

Chinese Education..........................................................................................225

US Education .................................................................................................228

European Education........................................................................................240

The Educational Response to Offshoring ...........................................................246


AN INTERNATIONAL PERSPECTIVE..........................................................254

US Policy.......................................................................................................254

Page 7

Australian Policy.............................................................................................262

Swedish Policy ...............................................................................................264

Indian Policy..................................................................................................267

Chinese Policy................................................................................................268



A Biographies of Job Migration Task Force Members..........................................275

B Selected Abbreviations and Acronyms ..........................................................285

Page 8

Globalization and Offshoring of Software

A Report of the ACM Job Migration Task Force

Executive Summary and Findings

Why this Study?

This study reports on the findings of a Task Force established by The Association for

Computing Machinery (ACM) to look at the issues surrounding the migration of jobs

worldwide within the computing and information technology field and industry. ACM initiated

this study to provide a deeper understanding of the trends in, and the forces behind, the

globalization and offshoring of software. Because ACM is an international educational and

scientific computing society, the study approached the issue of offshoring of software from

an international as opposed to a United States-centric perspective. Moreover, the task force

that conducted the study comprised not only computer scientists (ACM’s traditional

constituency) but also labor economists and social scientists from around the world. We

believe that this approach, and this perspective, are unique. Most reports on globalization

and offshoring are produced either by governments or national organizations, and thus

provide an inherently national perspective, or by consulting firms in pursuit of their own or

their clients’ business interests.

The primary purpose of the study is to provide ACM’s 83,000 members, the computing

field, the IT profession, and the public an objective perspective on current and future trends

in the globalization of the software industry so that ACM members can better prepare

themselves for a successful future in the system, software, and services portion of the

global information technology field. We also believe this extensive study will be of value to

those shaping the policies, priorities, and investments any country must make if it desires to

remain or become a part of the global software-systems-services industry.

Scope of the Study

This study reports on the current state of globalization and offshoring of software and

related information technology (IT) services. (Outsourcing refers to having work for a

company done by another organization. Offshoring refers to having this work done in

another country, whether or not it is done by part of the same company.)

The report is focused primarily on software systems work carried out in developing

countries for export, as opposed to work done in a developing country for their local market.

The ACM Task Force reviewed existing reports and data from around the world, and heard

in-person from many experts, on issues relevant to globalization and offshoring. In the

process, the Task Force took an in-depth look at the following:

1. The economic theories and data that underpin our current understanding of the forces

shaping globalization today and in the future.

2. Offshoring from the perspective of different countries—both developed and developing.

3. Offshoring from the perspective of different types of corporations.

4. The globalization of computing research.

5. The risks and exposure that offshoring engenders.

6. The implications for educational systems throughout the world.

Page 9

7. The political responses to the opportunities and disruptions that accompany


Each of these areas is explored in detail in a chapter of the report.

Findings and Recommendations

In reviewing many existing reports, data, theories, and perspectives, a number of key

findings and recommendations emerged.

1. Globalization of, and offshoring within, the software industry are deeply

connected and both will continue to grow. Key enablers of this growth are

information technology itself, the evolution of work and business processes,

education, and national policies.

The world has changed. Information technology is largely now a global field, business,

and industry. There are many factors contributing to this change, and much of this change

has occurred within the past five years. Offshoring is a symptom of the globalization of the

software-systems-services industry.

This rapid shift to a global software-systems-services industry in which offshoring is a

reality has been driven by advances and changes in four major areas:

1. Technology—including the wide availability of low-cost, high-bandwidth

telecommunications and the standardization of software platforms and business

software applications.

2. Work processes—including the digitalization of work and the reorganization of work

processes so that routine or commodity components can be outsourced.

3. Business models—including early-adopter champions of offshoring, venture capital

companies that insist the companies they finance use offshoring strategies to reduce

capital burn rate, and the rise of intermediary companies that help firms to offshore

their work.

4. Other drivers—including worldwide improvements in technical education, increased

movement of students and workers across national borders, lowering of national

trade barriers, and the end of the Cold War and the concomitant increase in the

number of countries participating in the world market.

2. Both anecdotal evidence and economic theory indicate that offshoring between

developed and developing countries can, as a whole, benefit both, but

competition is intensifying.

The economic theory of comparative advantage argues that if countries specialize in areas

where they have a comparative advantage and they freely trade goods and services over

the long run, all nations involved will gain greater wealth. As an example, the US and India

have deeply interconnected software industries. India benefits from generating new revenue

and creating high-value jobs; the US benefits from having US-based corporations achieve

better financial performance as a result of the cost savings associated with offshoring some

jobs and investing increased profits in growing business opportunities that create new jobs.

This theory is supported to some extent by data from the US Bureau of Labor Statistics

(BLS). According to BLS reports, despite a significant increase in offshoring over the past

five years, more IT jobs are available today in the US than at the height of the

boom. Moreover, IT jobs are predicted to be among the fastest-growing occupations over

the next decade.

Page 10

Some economists have recently argued that in certain situations offshoring can benefit

one country at the expense of another. While debate continues about this aspect of

theory/policy, the majority of the economic community continues to believe that free trade

is beneficial to all countries involved, though some argue that globalization may lead to

technology leaders’ losing their current dominant position.

In any event, economists agree that even if a nation as a whole gains from offshoring,

individuals and local communities can be harmed. One solution to this potential negative

impact is for corporations or their governments to provide programs that aid these

individuals and their related communities in once again becoming competitive. The cost of

such “safety-net” programs can be high and, thus, difficult to implement politically.

3. While offshoring will increase, determining the specifics of this increase is

difficult given the current quantity, quality, and objectivity of data available.

Skepticism is warranted regarding claims about the number of jobs to be

offshored and the projected growth of software industries in developing


Data for making good decisions about offshoring are difficult to obtain. Government data

as collected are not very helpful and do not adequately address the specific issue of

offshoring. The objectivity and quality of other data sources, especially the data in reports

from consulting firms and trade associations, is open to question, as these organizations

may be serving their own agendas. Projections are always more suspect than data on

current employment levels.

It is very difficult to determine how many jobs are being, or will be, lost due to offshoring.

The best data available are for the United States. Some reports suggest that 12 to 14

million jobs are vulnerable to offshoring over the next 15 years. This number is, at best, an

upper limit on the number of jobs at risk. To date, the annual job loss attributable to

offshoring is approximately 2 to 3 percent of the IT workforce. But this number is small

compared with the much higher level of job loss and creation that occurs every year in the

United States.

Thirty percent of the world’s largest 1000 firms are offshoring work, but there is a

significant variance between countries. This percentage is expected to increase, and an

increase in the amount of work offshored is consistent with the expected growth rate of 20

to 30 percent for the offshoring industries in India and China. Almost all estimates are

based on reports from national and international consulting firms and, thus, subject to


4. Standardized jobs are more easily moved from developed to developing

countries than are higher-skill jobs. These standardized jobs were the initial

focus of offshoring. Today, global competition in higher-end skills, such as

research, is increasing. These trends have implications for individuals,

companies, and countries.

The report considers several case studies of firms and how they are addressing offshoring,

including software service firms in low-wage nations and four types of firms in high-wage

nations: packaged software firms, software service firms, entrepreneurial start-up firms,

and established firms outside the IT sector. These cases show that the amount and diversity

of work being offshored is increasing; and companies, including start-ups, are learning how

to access and use higher skill levels in developing countries.

One example of a higher-skill area now subject to global competition is computing

research. Historically, the bulk of this research was carried out in only a few countries -

countries with high purchasing-power-parity adjusted gross domestic product (PPP GDP)

Page 11

and with a relatively large percentage of PPP GDP devoted to research and development.

This situation is changing rapidly and the trend looks inexorable. Many companies have

established research centers in multiple countries. Most of these companies retain strong

research operations in their home country. This fact, combined with increasing national

research investment in India and China, is leading to both an increase in the total worldwide

investment in research and a wider distribution of research activities around the world.

People are by far the most important asset in research. The historic advantage held by

Western Europe and the United States is not as strong today as in the past, given the

developments in the graduate education systems in China and India, increased

opportunities for research careers in those countries, and the rising national investment in

research. The United States, in particular, faces a challenge in its inability to recruit and

retain foreign students and researchers in the numbers it did in the past. Its dominance in

the research area is likely, therefore, to be challenged.

Finally, while there is no way of ensuring lifetime IT employment, there are steps that

students and IT workers can take to improve their chances of long-term employment in IT

occupations. These include obtaining a strong foundational education, learning the

technologies used in the global software industry, keeping skills up to date throughout their

career, developing good teamwork and communication skills, becoming familiar with other

cultures, and managing their careers so as to choose work in industries and jobs

occupations less likely to be automated or sent to a low-wage country.

5. Offshoring magnifies existing risks and creates new and often poorly

understood or addressed threats to national security, business property and

processes, and individuals’ privacy. While it is unlikely these risks will deter the

growth of offshoring, businesses and nations should employ strategies to

mitigate them.

When businesses offshore work, they increase not only their own business-related risks

(e.g., intellectual property theft, failures in longer supply chains, or complexity arising from

conflicting legal environments) they also increase risks to national security and individuals’

privacy. Businesses have a clear incentive to manage these new risks to suit their own

interests, but nations and individuals often have little awareness of the exposures created.

For example, many nations have adopted commercial off-the-shelf (COTS) software and

Internet Protocol technologies in IT-based military systems and critical infrastructure

systems. Many COTS systems are developed, in part or whole, offshore, making it

extremely difficult for buyers to understand all source and application code. This creates the

possibility that a hostile nation or non-governmental hostile agents (terrorist/criminal) can

compromise these systems. Individuals often are exposed to loss of privacy or identity

theft. Bank records, transaction records, call center traffic, and service centers all are being

offshored today. Voluminous medical records are being transferred offshore, read by

clinicians elsewhere, stored and manipulated in foreign repositories, and managed under

much less restrictive laws about privacy and security than in most developed countries.

These risks can be managed by companies and governments through the use of risk

mitigation strategies. For example, businesses should minimize access to databases by

offshore operations and encrypt data transmissions; offshoring providers should be vetted

carefully; companies should have security and data privacy plans and be certified to meet

certain standards; and service providers should not outsource work without the explicit

approval of the client. Nations can adopt stronger privacy policies, invest in research

methods to secure this data, or work on the development of nation-to-nation and

international treatment of both the data and how compromises will be handled.

Page 12

6. To stay competitive in a global IT environment and industry, countries must

adopt policies that foster innovation. To this end, policies that improve a

country’s ability to attract, educate, and retain the best IT talent are critical.

Educational policy and investment is at the core.

Building a foundation to foster the next generation of innovation and invention requires

Sustaining or strengthening technical training and education systems,

Sustaining or increasing investment in research and development, and

Establishing governmental policies that eliminate barriers to the free flow of talent.

Education is one of the primary means for both developed and developing countries to

mount a response to offshoring so their workforces can compete globally for IT jobs. In fact,

education has been a primary enabler of offshoring in the developing countries. India has

responded rapidly to the educational needs of its software export industry, especially

through its private universities and training organizations. China is addressing the

educational needs of its software industry through centralized planning.

There are, however, problems with both the Indian and Chinese educational systems.

India provides poor quality higher education outside its top tier of universities, the quality of

the faculty is uneven, research opportunities are not generally available to either students

or faculty, and there is a tension between providing a good education to a limited number of

people and providing access for all. The Chinese system is burdened with an emphasis on

rote learning, a reward system for faculty that has not yet been transformed fully to reward

research by faculty and their students, and problems moving from a central planning to a

competitive funding system that rewards merit and entrepreneurship.

Developed nations can use education as a response to offshoring in order to protect

national interests. It can, however, be complex for a nation to address offshoring through

education for several reasons: educational systems are complex, with multiple degrees and

multiple majors preparing one for an IT career; the nature of the software work that is

being offshored is changing rapidly; it is difficult to forecast national supply and demand

needs for software workers; governments can only indirectly affect supply and demand in

many nations; and it is difficult to translate an educational response to offshoring into

practical curriculum reform. For example, the United States educational system is still trying

to understand how to change its curriculum to address application domain knowledge, a

global workplace, and maintaining its innovative edge. In addition, the United States faces

long-term challenges from falling interest and skills in math and science programs in its

primary education system. The European Union is struggling with the implementation of the

Bologna Directive to achieve a single European educational framework.

There are some general principles that all countries can follow to mount an effective

educational response to offshoring:

1. Evolve computing curriculum at a pace and in a way that better embraces the

changing nature of IT.

2. Ensure computing curriculum prepare students for the global economy.

3. Teach students to be innovative and creative.

4. Evolve curriculum to achieve a better balance between foundational knowledge of

computing on the one hand, and business and application domain knowledge on the


5. Invest to ensure the educational system has good technology, good curriculum, and

good teachers.

Page 13


Globalization of, and offshoring within, the software industry will continue and, in fact,

increase. This increase will be fueled by information technology itself as well as government

action and economic factors and will result in more global competition in both lower-end

software skills and higher-end endeavors such as research. Current data and economic

theory suggest that despite offshoring, career opportunities in IT will remain strong in the

countries where they have been strong in the past even as they grow in the countries that

are targets of offshoring. The future, however, is one in which the individual will be situated

in a more global competition. The brightness of the future for individuals, companies, or

countries is centered on their ability to invest in building the foundations that foster

innovation and invention.

Page 14


In the spring of 2004 we were asked by the Association of Computing Machinery (ACM) to

chair a task force to study the phenomenon of information-technology offshoring. Offshoring

was a hot topic in early 2004. Since the dot-com and telecommunication crashes of the

early 2000s, offshoring appears to be the proverbial "third shoe" to hit the IT sector in the

United States. While articles on offshoring and outsourcing appeared in the media weekly,

sorting out facts from fiction was exceedingly difficult. While it was clear that offshoring was

a boon to providers in developing countries, debates raged on its impact on developed

countries. Getting a clear, factual picture of IT offshoring was undeniably important and


ACM offered us a free reign in conducting this study, subject to two constraints. First, the

study had to look at offshoring from a global perspective, reflecting ACM's position as an

international organization. Previous studies of offshoring have typically taken a national,

usually American, perspective. Second, the study had to be completed roughly within one

year, which implied that it had to be a secondary study, based on published material, rather

than a primary study, doing its own collection of data. Early on we decided to focus on the

software side of IT. Offshoring of IT manufacturing has been going on for a number of

years; the phenomenon that took off during the early 2000s was the offshoring of software.

Our hope is this report sheds much-needed light on software offshoring. It points out that

offshoring is a symptom of globalization, which has been an inexorable economic force since

1990, while examining the specific forces that drive software offshoring, both at the country

level and at the firm level. It surveys the debate on the economic impact of offshoring, and

examines the available data, pointing out the paucity of reliable relevant data. The report

also shows how IT research has been leading the offshoring trend. It highlights risks and

exposures to individuals, corporations, and countries created or magnified by offshoring.

Finally, it portrays the opportunities and challenges that offshoring poses to IT education in

both developing and developed countries.

The Task Force was assembled during the second half of 2004. Bill Aspray, who has

experience with work force studies, agreed to serve as executive consultant and primary

editor for this study; indeed, it could not have been carried out without him. To ensure a

broad perspective, we recruited around 30 Task-Force members, computer scientists,

economists and sociologists from the US, Europe, Israel, India and Japan (see listing below

and biographies). We are grateful to all of them for volunteering their time and efforts.

The Task Force held four meetings: in Chicago, IL, Oct. 8-9, 2004; Washington, DC, Dec.

3-4, 2004; Palo Alto, CA, March 4-5, 2005; and New York, NY, May 13, 2005. During it first

meeting, the Task Force scoped the study, decided what the main topics should be, and

divided into several committees, with some members serving on more than one. Roughly,

each committee was focused on one topic, which is covered by one chapter in the final

report. The next two meetings were dedicated to hearing perspectives by many experts and

scholars (listed below), while committees continued their work during and between the

meetings. At the final meeting, the committees presented drafts of their reports and

received feedback from the rest of the Task Force.

The committees prepared the final drafts of their reports during the summer of 2005.

These drafts went then to Aspray for editing. The edited versions were then sent to

reviewers; each chapter was vetted by several reviewers. The process of review and

revision continued through the fall of 2005.

During its work, the Task Force has reviewed hundreds of articles on the subject of

offshoring, and IT offshoring in particular. To aid the Task Force in its work, Aspray

Page 15

prepared an extensive annotated bibliography. While this bibliography is too extensive to be

included in the printed report, ACM is making it available online

( as a service to its members.

We appreciate the efforts of the many individuals who helped the Task Force carry out its


Frank Mayadas and Moshe Y. Vardi

Task Force Co-Chairs, December 2005.


The following people gave generously of their time to serve on the ACM Job Migration

Task Force, which produced this report. Biographies of the Task Force members are given

in the appendix.

Task Force Executive Committee:

Frank Mayadas (Sloan Foundation), co-chair

Moshe Y. Vardi (Rice University), co-chair

William Aspray (Indiana University), executive consultant

John White (ACM), ex officio

Task Force Members:

Alok Aggarwal (Evalueserve)

Frances E. Allen (IBM)

Stephen J. Andriole (Villanova University)

Ashish Arora (Carnegie Mellon University)

G. Balatchandirane (University of Delhi)

Burt S. Barnow (Johns Hopkins University)

Orna Berry (Gemini Israel Funds, Adamind, and Prime Sense)

Michael Blasgen (consultant)

Arndt Bode (Technische Universität München)

Jean Camp (Indiana University)

Seymour E. Goodman (Georgia Institute of Technology)

Vijay Gurbaxani (University of California, Irvine)

Juris Hartmanis (Cornell University)

Page 16

Charles House (Intel)

William Jack (SAIC)

Martin Kenney (University of California, Davis)

Stefanie Ann Lenway (University of Illinois at Chicago)

Vivek Mansingh (Dell India Development Center)

Göran Marklund (Swedish Offices of Science and Technology, Washington, DC)

Peter Mertens (University of Erlangen-Nuremberg)

Rob Ramer (TerraFirma Security)

Bobby Schnabel (University of Colorado, Boulder)

Bankim Shah (BRS Associates)

Marie Stella (US Federal Aviation Administration)

Valerie E. Taylor (Texas A&M University)

Takashi Umezawa (Kokushikan University)

Roli Varma (University of New Mexico)

Richard C. Waters (Mitsubishi Electric Research Laboratories)

Stuart Zweben (Ohio State University)

We thank the graduate students who provided research assistance to the Task Force:

Alla Genkina (Indiana University and University of California, Los Angeles)

Matthew Hottell (Indiana University)

We thank the staff of the ACM who supported this project:

Marcia Boalen

Mark Mandelbaum

David Padgham

Cameron Wilson

We appreciate the people who came to make presentations to the Task Force in

Washington, DC or Palo Alto, CA:

Josh Bivens (Economic Policy Institute)

Rafiq Dossani (Stanford University)

Ralph Gomory (Sloan Foundation)

Dan Griswold (Cato Institute)

Page 17

Ron Hira (Rochester Institute of Technology)

Lori Kletzer (University of California, Santa Cruz)

Leonard Liu (Augmentum)

Catherine Mann (Institute for International Economics)

Anita Manwani (Agilent Technologies)

Carol Ann Meares (US Commerce Department)

Peter Neumann (SRI)

Suhas Patil (Cirrus Logic)

John Sargent (US Commerce Department)

Loren Yager (US General Accountability Office)

We also thank the many professionals who reviewed a preliminary draft of sections of the


Rakesh Basant (Indian Institute of Management Ahmedabad)

Allan Borodin (University of Toronto)

Erran Carmel (American University)

Lorrie Cranor (Carnegie Mellon University)

Anthony D'Costa (University of Washington)

Dorothy Denning (Naval Postgraduate School)

Peter Denning (Naval Postgraduate School)

Rafiq Dossani (Stanford University)

Ben Fried (Morgan Stanley)

Peter Hart (Ricoh Innovations)

Ron Hira (Rochester Institute of Technology)

John King (University of Michigan)

Randy Kirihara (Target)

Maria Klawe (Princeton University)

Lori Kletzer (University of California Santa Cruz)

Rich LeBlanc (Southern Catholic College)

Peter Lee (Carnegie Mellon University)

Thomas Lookabaugh (University of Colorado Boulder)

Catherine Mann (The Institute for International Economics)

Andrew McGettrick (University of Strathclyde)

J Strother Moore (University of Texas)

Sharon O'Bryan (O'Bryan Advisory Services Inc.)

Page 18

David Patterson (University of California Berkeley)

Eric Roberts (Stanford University)

Mari Sako (The Saïd Business School Oxford University)

Kevin Schofield (Microsoft)

Alan Selman (University at Buffalo, The State University of New York)

Manuel Serapio (University of Colorado, Denver)

Russ Shackelford (Stanford University)

Eugene Spafford (Purdue University)

Elizabeth Sparrow (British Computer Society)

Manuel Trajtenberg (Tel-Aviv University)

N.V. "Tiger" Tyagarajan (GECIS)

Robert A. Walker (Kent State University)

David Waltz (Columbia University)

Tim Wedding (Government Accountability Office)

Loren Yager (US Government Accountability Office)

Yoram Yahav (Technion Institute of Management)

Andrew Yao (Tsinghua University)

Page 19

Globalization and Offshoring of Software

A Report of the ACM Job Migration Task Force


1. The Big Picture

Over the past decade, low-wage countries such as India have developed vibrant, exportoriented

software and IT service industries. Attracted by available talent, quality work, and

most of all low cost, companies in high-wage countries, such as the United States and the

United Kingdom, are increasingly offshoring software and service work to these low-wage

countries. Trade (together with automation) cost many jobs in the manufacturing sector to

be lost from the West and many developing nations in East Asia to increase their wealth and

industrial prowess since 1970. Changes in technology, work organization, educational

systems, and many other factors have caused service work—previously regarded as

immune to these forces—also to become tradable. This trade in services, led by the trade in

software and IT-enabled services, presents many opportunities and challenges for

individuals, firms, and policymakers in both developed and developing nations.

Many people in the United States and Western Europe fear that sending software work

offshore will cause wage and job suppression in the high-wage countries. Others believe

that the process of getting good labor at lower prices will make the economy more

productive, enabling the creation of new wealth and new jobs. Many people in the low-wage

countries are excited by the economic development that their software and service

industries are bringing them; while some are concerned about the side effects such as

congestion, pollution, and loss of traditional cultural values. One thing that is clear is that

the globalization of software is here to stay, so that policymakers, educators, and employers

all need to address the realities of offshoring. This includes, for example, how to help people

whose jobs are shipped to another country to get assistance with their careers, how to

create innovative environments that help to create new jobs, and how to revamp

educational systems for the realities of a globalized world.

“Offshoring” is the term used here. It is a term that applies best to the United States

because, even though the United States does outsource work to Canada and Mexico, most

of its work is sent over the seas—mostly to India, but also to China, Malaysia, the

Philippines, and many other places. Germany, for example, also sends work across its

borders, including to Eastern Europe, but there is no water—no shore—to cross. Some of

the work that is offshored is sent to entrepreneurial firms established in these low-wage

countries. Other times, multinationals headquartered in high-wage countries open

subsidiaries in the low-wage countries to work on products and services for their world

market. Multinationals may also open facilities in these low-wage countries in order to

better serve the local market there, but that situation is not the primary interest of this


There are at least six kinds of work sent offshore related to software and information

technology: (1) programming, software testing, and software maintenance; (2) IT research

and development; (3) high-end jobs such as software architecture, product design, project

management, IT consulting, and business strategy; (4) physical product

manufacturing—semiconductors, computer components, computers; (5) business process

outsourcing/IT Enabled Services—insurance claim processing, medical billing, accounting,

bookkeeping, medical transcription, digitization of engineering drawings, desktop publishing,

Page 20

and high-end IT enabled services such as financial analysis and reading of X-rays; and (6)

call centers and telemarketing. Our primary interest is with the first three of these

categories, which are the ones most closely associated with the transfer of software work

across national boundaries. However, it is almost impossible to study offshoring without at

least at times considering the other three categories of work as well. This is because

companies that do one of these kinds of software work may also do several other kinds of

offshore work as part of their product and service line of offerings; and companies that send

work offshore may send work of several kinds. Because companies and industries

intermingle these categories of work, so does most statistical data that tracks this

industry—and it is often impossible to disaggregate data to capture information about only

the categories of work of greatest concern here. Thus we focus on the first three categories

but discuss the others in passing.

The countries that send work offshore are primarily developed nations. The United States

followed by the United Kingdom have been the largest offshorers, but other countries in

Western Europe, Japan, Korea, Australia, and even India send work offshore. The countries

that do the work fall into four categories: (1) those that have a large capacity of highly

educated workers and have a low wage scale (e.g., India, China); (2) those that have

special language skills (e.g., the Philippines can serve the English and Spanish customer

service needs of the United States by being bilingual in these languages); (3) those that

have geographic proximity (“nearsourcing”), familiarity with the work language and

customs, and relatively low wages compared to the country sending the work (e.g. Canada

accepting work from the United States, the Czech Republic accepting work from Germany);

and (4) special high-end skills (e.g., Israeli strength in security and anti-virus software).

There are many drivers and enablers of offshoring. These include: (1) The dot-com boom

years witnessed a rapid expansion of the telecommunications system, making ample, lowcost

broadband available in many countries at attractive rates. This made it possible to

readily transfer the data and work products of software offshoring. (2) Software platforms

were stabilized, with most large companies using a few standard choices: IBM or Oracle for

database management, SAP for supply chain management, and so on. This enabled

offshoring suppliers to focus on acquiring only these few technologies and the people who

are knowledgeable about them. (3) Companies are able to use inexpensive commodity

software packages instead of customized software, leading to some of the same

standardization advantages as with software platforms. (4) The pace of technological

change was sufficiently rapid and software investments became obsolescent so quickly that

many companies chose to outsource IT rather than invest in technology and people that

would soon have to be replaced or retrained. (5) Companies felt a competitive need to

offshore as their competition began to do so. (6) Influential members from industry, such as

Jack Welch from General Electric, became champions of offshoring. (7) Venture capitalists

pushed entrepreneurial startups to use offshoring as a means to reduce the burn rate of

capital. (8) New firms emerged to serve as intermediaries, to make it easier for small and

medium-sized firms to send their work offshore. (9) Work processes were digitalized, made

routine, and broken into separable tasks by skill set—some of which were easy to

outsource. (10) Education became more globally available with model curricula provided by

the professional computing societies, low capital barriers to establishing computer

laboratories in the era of personal computers and package software, national plans to build

up undergraduate education as a competitive advantage, and access to Western graduate

education as immigration restrictions were eased. (11) Citizens of India and China, who had

gone to the United States or Western Europe for their graduate education and remained

there to work, began to return home in larger numbers, creating a reverse Diaspora that

provided highly educated and experienced workers and managers to these countries. (12)

India has a large population familiar with the English language, the language of global

business and law. (13) India has accounting and legal systems that were similar to those in

Page 21

the United Kingdom and the United States. (14) Global trade is becoming more prevalent,

with individual countries such as India and China liberalizing their economies, the fall of

Communism lowering trade barriers, and many more countries participating in international

trade organizations.

At first it was believed that the only software work that would be offshored was low-level

work, such as routine software maintenance and testing, routine business office processes,

and call centers. Offshoring suppliers, however, have made strong efforts to move up the

value chain and provide services that have higher value added because this is where there

is the greatest opportunity for profit. Research and development, project integration, and

knowledge process outsourcing such as reading X-rays and doing patent checking are

increasingly being offshored. Today, some people believe that any kind of software or ITenabled

work can be offshored. While there is an element of truth in this belief, there are

some important caveats. Some kinds of work have not been offshored. Even if it is possible

to offshore a particular type of work, it does not mean that every job of that type actually

will be offshored. In fact, there are a number of reasons why a company might not wish to

offshore work: (1) the job process has not been made routine. (2) The job cannot be done

at a distance. (3) The infrastructure is too weak in the vendor country. (4) The offshoring

impacts too negatively on the client firm such as the client firm losing control over an

important work element, losing all its in-house expertise in an area, or too high a loss of

worker morale in the client firm. (5) Risks to privacy, data security, or intellectual property

are too high. (6) There are not workers in the supplier firm with the requisite knowledge to

do the job, which happens for example when the job requires application domain knowledge

as well as IT knowledge. (7) Costs of opening or maintaining the offshore operation are too

expensive. (8) There are cultural issues that stand between the client and vendor. (9) The

company can achieve its goal in another way, such as outsourcing within its home country

or consolidating business operations.

One might wonder whether IT is still a good career choice for students and workers in

countries that offshore software and IT services work. Despite all the publicity in the United

States about jobs being lost to India and China, the size of the IT employment market in

the United States today is higher than it was at the height of the dot-com boom.

Information technology appears as though it will be a growth area at least for the coming

decade, and the US government projects that several IT occupations will be among the

fastest growing occupations during this time. There are some things that students and

workers in this field should do to prepare themselves for the globalized workplace. They

should get a good education that will serve as a firm grounding for understanding the

rapidly changing field of IT. They should expect to participate in life-long learning. They

should hone their “soft skills” involving communication, management, and teamwork. They

should become familiar with an application domain, especially in a growth field such as

health care, and not just learn core technical computing skills. They should learn about the

technologies and management issues that underlie the globalization of software, such as

standard technology platforms, methods for re-using software, and tools and methods for

distributed work.

2. The Economics of Offshoring

Much of the economic debate about offshoring centers around whether the theory of

comparative advantage applies to the offshoring of software and IT services. Economists

have argued on both sides of the issue. The arguments are sophisticated and nuanced, and

the results often depend on whether the underlying assumptions hold in the current context.

While a majority of economists are proponents of free trade, the underlying question is an

empirical one and can be answered by analyzing reliable data when it becomes available.

Page 22

The theory of comparative advantage states that if each country specializes in the

production of goods where it has a comparative (relative) advantage, both countries can

enjoy greater total consumption and well being in aggregate by trading with each other.

Offshoring enables, for example, US firms to lower costs and save scarce resources for

activities in which they have a relative advantage, while offshoring has led to significant

employment and wage gains for Indian workers and rapid profit and revenue increases for

Indian businesses.

What the theory of comparative advantage does not mean is that all members of society

will benefit from trade. In general, imports of an “input” have economic effects that are

similar to those of an increase in the supply of the input, namely, lower returns to the

suppliers of the input, lower costs of production, and lower prices for consumers. If the

input were a service, the wages and salaries of those producing the service would fall, but

so also would costs for firms that are buyers of the service. In the exporting country, the

opposite effects hold. That is, the returns to the owners or suppliers of the service or input

increase and the wages of the employees at the service providers increase due to the higher


Economists believe that trade generally leads to significant gains for society. These gains

are not inconsistent with employment losses in specific sectors that will cause economic

pain to the workers affected. To achieve an equitable result, many analysts believe that it is

important to establish a safety net that provides income and training opportunities to

affected workers. Components of the safety net might include extended unemployment

benefits, wage insurance, and retraining.

A key assumption underlying the theory of comparative advantage is that the economy

enjoys full employment. Thus, this theory is best thought of as a theory of the long-term, in

which workers displaced by imports or offshoring find work in other sectors. By contrast,

most popular discussions of the offshoring phenomenon tend to focus on questions such as

“where will the new jobs be created” and “can the workers be retrained for these new jobs”.

In general, peering into the crystal ball to predict where and what types of new jobs will be

created is both difficult and unrewarding. A dynamic economy such as that of the United

States creates and destroys millions of new jobs in response to changes in tastes, and more

importantly in response to innovations and advances in technology. There is no guarantee

that the economy will continue to create these new jobs, but policy makers can take some

comfort from the historical evidence that thus far it has managed to do so. The key to job

creation is of course the ability of the economy to rapidly generate and adopt

innovations—new types of goods and services, and productivity-enhancing process


In general, trade stimulates innovation and economic growth in both trading partners.

Some, such as Ralph Gomory and Gregory Baumol, have argued that innovation

opportunities create new possible conflicts of interest between trading partners. For

example, insofar as offshoring stimulates, in countries such as China, innovation and

productivity growth in goods and services where developed countries such as the United

States enjoy a comparative advantage, this will cause the “terms of trade” to become less

favorable over time for the United States. In other words, even if free trade is the best

policy, it may well be that free trade, by stimulating innovation overseas, may impose longterm

losses. However, Gomory and Baumol’s analysis shows that this conflict of interest is

present when the two trading partners are at similar stages of development. Since most

offshoring involves countries at very different levels of development, this conflict of interest

is presently unlikely.

In the IT services sector, there is a related concern. Currently, it is efficient to offshore

“low-end” IT services, such as coding or maintenance, to a low-wage country while “highPage


end” activities, such as requirements analysis, design, and R&D, remain in the high-wage

country. The concern is, however, that eventually the high-end IT activities would also

move offshore. Were this to happen, the current technology leaders (United States,

Germany, Japan, United Kingdom, et al.) may relinquish that leadership role. There is some

anecdotal evidence that some IT process innovations are moving to low-wage, offshoring


Most economists, however, argue that current technology leaders will not lose their

technological leadership position. Even if production moves to other countries, history shows

that in many industries the locus of production and the locus of invention are physically

separated. There are two key resources required to remain at the center of innovation in

software: access to talented designers, software engineers, and programmers; and

proximity to a number of large and technically sophisticated users. Current technology

leaders, and the United States in particular, currently dominate on both counts. More

broadly, the United States has other important capabilities, including the best universities

and research institutions, highly efficient capital markets, flexible labor markets, the largest

consumer market, business-friendly immigration laws, and a large and deep managerial

talent pool. As a result, the evolution of business in the United States has followed a

consistent pattern: launch innovative businesses here, grow the business, and as products

and services mature migrate lower-value-added components and intermediate services over

time to lower-cost countries. Nevertheless, there are those who argue that globalization will

diminish the comparative advantage of current technology leaders, which may lead to the

loss of their current dominant position and create a long period of adjustment for their


Data on current and future trends of offshoring leave much to be desired. First, the

definitions of offshoring vary from one study to another, making it hard to compare

statistics. For example, some studies count all service jobs, some count IT jobs, some

include IT-enabled jobs, and some are simply not precise about what they are counting.

Second, there is a question of what metric to use in measuring the extent and trends in

offshoring. One might measure, for example, jobs lost in the developed country, jobs in the

developing country’s IT industry, or dollar value of business outsourced. In the case of each

of these metrics, however, it is either difficult to make the measurement or the metric is not

directly enough relevant to the offshoring situation. For example, it is difficult to calculate

dollar value of business offshored because these are internal transfer costs for

multinationals, which they may not be willing to report or do not report in an appropriately

disaggregated way.

Projections of future trends are more suspect than data on the current situation. One type

of projection identifies types of jobs that are vulnerable to offshoring. These vulnerability

projections provide at best a high upper bound on expected job loss, and for this reason

they are blunt policy-making tools. It may be that routine programming jobs are vulnerable

to offshoring, but it is highly unlikely that every last one of them will be lost to offshoring.

Moreover, even in cases where the methodology is sound and soundly applied, projections

of any kind about the future are much less likely to be accurate than data about today’s or

yesterday’s situation since it is difficult to predict all the factors that will come into effect

over time.

Another important issue to consider is the source of the data. Data from the United States

and many other national governments tends in general to be reliable. The US government,

however, collects data to handle established policy issues. If a new phenomenon arises, the

existing data sets may not be well suited to studying the new policy issue. This is the case

with offshoring. US data on job layoffs and on service trade are both designed for other

purposes, and there is widespread belief among economists that both seriously undercount

offshoring trends. Data collected and analyzed by trade associations and consulting firms

Page 24

may be very useful, but there is skepticism in the economic community about the quality of

these data in many cases because the methods for collecting and analyzing the data are

often not made available for scrutiny, the data they collect (from members of their

organization) may not be a representative sample of society as a whole, and these

organizations have particular objectives in mind that they hope their data will bolster.

The United States is the source of the greatest number of offshored jobs and where the

phenomenon has received the greatest attention. But even for the US, it is difficult to be

certain of the extent of offshoring. Federal data is not very helpful, and most of the existing

data comes from consulting firms. The numbers generally indicate that 12 to 14 million jobs

in the United States are vulnerable to relocation through offshoring, and that annual losses

have ranged from under 200,000 to about 300,000 service jobs from the United States to

offshoring. The number of IT jobs is somewhat lower than these estimates because these

estimates include service jobs such as working in call centers and sometimes other ITenabled

services such as business process and knowledge process offshoring. Importantly,

these estimates do not include newly created jobs. The consensus seems to be that about

20% of US companies are currently offshoring work but that the percentage is rising. The

current value of offshoring contracts from the United States seems to be in the $10 to 20

billion range, with an expectation of rapid growth. It should be remembered, however, that

we do not know the methods used to arrive at these numbers and how independent the

data from one consulting firm’s study is from that of another.

Statistics for the entire world or for other individual countries are even harder to come by

and more suspect than those for the United States. The annual dollar value of worldwide

offshoring trade for recent years has been estimated to be between $1.3 billion and $32

billion, depending on whether certain exported products are counted and whether the

numbers for multinational companies are included. An estimated 30% of the world’s largest

1000 firms are offshoring work. Europe has lower levels of offshoring than the United

States. It is estimated that only 5% of European businesses (of all sizes) are offshoring, and

at most 2 to 3% of European IT workers will lose their jobs to offshoring by 2015. The

United Kingdom has the highest rate of work sent offshore of any European nation, with an

estimated 61% of firms now offshoring. In Germany, only 15% of companies are now

offshoring, and perhaps a total of 50,000 German jobs have been lost to offshoring so far;

however, there seems to be an increase in German offshoring in the recent past. Statistics

about India show a vibrant IT industry, with annual growth of 20 to 30%, the vast majority

of the growth coming in the export rather than the domestic market. Data on the rest of the

world are too spotty to trust.

3. Understanding Offshoring from a National Perspective

The first countries to develop software industries primarily for export rather than domestic

purposes were Ireland and Israel. The big player to come in a little later was India,

beginning in the mid-1970s and growing rapidly from the late 1990s. To some degree, a

global division of labor is beginning to form: India serving the English-speaking world,

Eastern Europe and Russia serving Western Europe, and China serving Japan. But India is

also providing service to Western Europe, and China provides service to the United States.

In addition, there are many smaller supplier countries. The greatest attention is given in

this report to the United States and India, the two biggest players.

The United States has historically dominated and continues to dominate the software and

services industry, with about 80% of global revenue. It is highly dominant in the packaged

services industry, with 16 of the top 20 companies worldwide, and slightly less commanding

but still dominant in the software services sector, with 11 of the top 20 companies. This

dominance is due to a number of factors, including a legacy of government funding of R&D,

Page 25

computer science research in the open US higher education system, early adoption by

sophisticated users, the world’s largest economy and market, and leading semiconductor

and data storage industries that helped to spread the use of computing.

The centrality and dominance of the US industry has been a given during the past five

decades. What is emerging is the globalization of the software and software services

industry. This creates opportunities around the world for people and companies in both

developed and developing countries to participate in this profitable industry. It also creates

challenges for the former leaders, notably the United States, Western Europe, and Japan.

Software services is India’s largest export. As a large developing nation, India faces many

challenges, including high rates of poverty, corruption, and illiteracy; a substandard

infrastructure; excess government regulation; and various other problems typical of a poor

nation. These challenges are offset by a number of strengths, especially for software and

services production. It has a long history of developing capable mathematicians. India is

unique because of the large number of individuals with adequate English language

capability, and also for the large cadre of Indian managerial and technical professionals

working in North American and, to a lesser degree, in European high-technology

occupations and organizations. For those who can afford it, India has a strong and highly

competitive K-12 educational system emphasizing science and mathematics. Despite its

democratic socialist tradition that involved large amounts of bureaucracy and state

regulation, it has been a market economy and has a history of managerial education and

competence. These assets have given India many advantages in establishing a software

export industry.

India’s software export industry began in 1974, when it began sending programmers to

the United States to do work for the Burroughs Corporation. Political liberalizations related

to trade in the 1970s and again in the early 1990s helped to support the development of the

Indian software industry. Offering solutions to the Y2K problem helped the industry to grow

substantially. The industry expanded beginning in the late 1990s, first by bringing back to

India much of the software development, maintenance, and testing work it had previously

done on the client’s premises, then developing export businesses in business process

offshoring, call centers, and research and development. India is moving up the value chain

and is seeking people with considerably more skill than low-level programmers to do these

higher value jobs. Software and service export firms in India are growing at 20 to 25% per

year according to the best statistics available, and each of the three leading Indian software

firms (Infosys, TCS, and Wipro) already employs over 40,000 people.

India is likely to continue to grow its software industry in scale, scope, and value-added.

There is little reason to believe that offshoring as a process will end in the foreseeable

future, but it could slow down. The enormous investment by leading software multinationals

will expand the number of Indian project managers with strong managerial skills. This,

together with the relocation of portions of startup firms to India, is likely to result in greater

levels of entrepreneurship and lead to firms able to sell their skills on the global market. The

offshoring of IT services and software for export will dominate the near future of the Indian

software industry. There are several possible trajectories. Custom projects could become

more complex and large, leading Indian software professionals to move from programming

into systems integration and systems specification and design. The average size of projects

Indian firms are undertaking has already grown from 5 person-years in 1991 to 20 personyears

in 2003. As multinationals deepen their Indian operations, domain skills are

developing in India and some other nations, so that managed services are likely to become

more important; this will match global trends in the outsourcing of applications

management and business processes.

Page 26

Despite the fact that India’s software production for the US market exceeds that of any

other nation, it holds only a small share of the global market for all software value-added.

The only part of the software value chain in which India has made substantial inroads is in

applications development, where it has captured 16.4 percent of the world market. But

applications development is only approximately 5 percent of the entire global software

services market. This implies that there is much room for growth. In order to grow, the

Indian industry will have to shift to more complex activities by securing larger projects,

undertaking engineering services, integrating and managing services, or bidding on projects

that include transforming a client’s entire work process. India, however, will have some

difficulty achieving this growth unless it strengthens its R&D capability.

Software offshoring to India is likely to grow, not only through the continued growth of

indigenous Indian firms, but also because foreign software firms are increasing their

employment in India in product development and particularly in software services.

Competition is likely to grow between multinationals based in developed countries, such as

Accenture, IBM, and Siemens Business Services, and the large Indian firms, such as HCL,

Infosys, TCS, and Wipro, as the Indian companies expand their global reach and the

multinationals expand their operations in low-cost countries. The Indian subsidiaries of

multinationals play an important role in the development of India’s software capabilities,

because they are more willing to undertake high value-added activities, such as software

product development, within their own subsidiary in India than they are to send the work to

an Indian independent firm.

For at least the medium term, India should be able to retain its position of primacy for

software offshoring from the English-language world. In the longer term, unless India

makes an even greater effort to upgrade its universities and the technical capabilities of its

graduates, China may become an important alternative destination.

China’s software and services industry does not currently have a major impact on the

world economy. The industry is highly fragmented into many small companies, few of which

are large enough to take on large projects for developed nations. The hardware industry is

well established in China, and in the future it may drive the software industry to a focus on

embedded software. Unlike India, where the multinationals are focused mainly on serving

the world market, in China multinationals are more focused on positioning themselves to

serve the enormous, emerging domestic Chinese market.

Japan has the second largest software and services industry in the world, after the United

States; and it is the fastest growing industry in Japan. Japan makes games software and

custom software for the world market and packaged software for its domestic market. It

imports a significant amount of systems and applications software from the United States;

and it calls on China and India to provide custom software.

There are three typical patterns of Japanese offshoring. Most commonly, a Japanese firm

will identify a need for custom software, contract with a Japanese IT company to provide

the software, and the IT company will in turn contract with a Japanese subsidiary of a

Chinese firm to do the programming work. This programming used to be done almost

exclusively in Japan, but as the cost of locating Chinese workers in Japan has become

expensive, more and more of the programming is being done in China. A second approach

that is more recent is for Japanese firms to invest in China to form wholly owned

subsidiaries or joint ventures with Chinese firms. A third approach is for multinational

corporations to move programming and back-office functions of their Japanese subsidiaries

to lower-cost locations, often in China. The Dalian software park in China is growing rapidly

as a result of this emerging Japanese business. The amount of offshoring from Japan is still

small, but cost pressures are likely to cause it to increase; and since Japan has such a large

software industry, the opportunities for offshoring are considerable.

Page 27

The European Union represents the second largest market in the world for software and

IT services, after the United States. There are many differences, however, from country to

country, and the European Union cannot be viewed as a unified, homogeneous market. The

European software industry and employment patterns are different from those of the United

States, with much more software production done in-house and embedded in physical

products. This does not prevent offshoring, and certainly many leading European industrial

firms are establishing offshore facilities to produce embedded software. Much of this

employment is subsumed under R&D and other activities such as application-specific

integrated circuit design.

About two-thirds of the work offshored from Europe is offshored by the United Kingdom.

Continental European firms continue to lag UK firms in sending software work across their

borders. The Germanic and Nordic nations have only recently begun to build offshore

software and software service delivery capabilities, but firms with global practices such as

SAP, Siemens, and others are moving rapidly to build their offshore capabilities in Eastern

Europe, China, and India. The geography of European offshoring will be somewhat different

from that of the United States in that Nordic and Germanic firms will use Eastern Europe

and Russia in addition to India. Latin (Romance-language-speaking) Europe has been slower

to begin offshoring, but now its major firms are sending work to Romania, Francophone

Africa (particularly Morocco), and Latin America in addition to India. Despite these

geographical differences, there is no reason to believe that the pressures to offshore

software-related work will be substantially different than in the Anglophone nations. In part

this is because the US-based multinationals with strong global delivery capabilities, such as

Accenture, EDS, Hewlett-Packard, and IBM, are present and competitive in all European

markets. European firms may continue to experience a lag due to union and government

opposition to offshoring, but their cost and delivery pressures are similar to those

experienced by US firms.

In Russia, software was a relatively neglected field during the Soviet era, but in the 1990s

as the country transitioned to a market economy, many scientists and engineers moved

from low-paid government and university positions into entrepreneurial firms and Russian

subsidiaries of multinationals; and some of these people entered the software field. So far

there are relatively few programmers. Wages are low. Technical skill level is high, but there

is little project management experience. Software firms are typically small, not able to take

on large international software integration projects. Nevertheless, the high skill level of the

Russian research community, a legacy of its Soviet history, has led Intel and a few other

multinationals including Boeing, Motorola, Nortel, and Sun to open R&D facilities in Russia.

4. Understanding Offshoring from a Company Perspective

Instead of examining offshoring by country, it is also possible to examine offshoring by

the type of company. We will consider five types of firms. The first are large, established

software firms headquartered in developed nations that make and sell packaged software.

Examples include Adobe, Microsoft, and Oracle. As a general rule, the largest and most

successful packaged software firms are headquartered in the United States; the notable

exception is SAP in Germany.

Most large packaged software firms have global operations. In many cases, their offshore

operations are for localization work for the local domestic market. However, particularly in

the case of India, and also to some degree in Russia, the work is for development of their

worldwide software packages. Locating in these low-wage countries enables these firms to

have access to lower-cost programmers, many of whom are comparable in skill levels to the

company’s workers in the developed nations. This is not the only benefit. Having operations

in other time zones can speed up production by facilitating round-the-clock production.

Page 28

These opportunities are encouraging major packaged software firms to expand their

workforce in India and other lower-cost nations.

Offshoring will have a complicated effect on the packaged software firms. First, it might

and likely will put employment pressure on developed nation software firms to decrease

employment in the developed nations. On the other hand, the lower cost and faster

production could allow the development of new features in old software and could contribute

to the creation of new, well-priced software products, which would in turn increase income

for these firms and perhaps lead to greater hiring.

Next we consider large, established software firms headquartered in developed nations

that are large providers of software services. These companies may also provide packaged

software, though not all of them do so. Examples include Accenture, EDS, and IBM.

Software service firms have been among the fastest growing firms in the IT sector, and in

general they are far larger than the packaged software firms. Firms coming from the

software side (e.g., Hewlett Packard or IBM) and from the service side (e.g., Accenture) are

converging. In the case of IBM, this has been through both direct hiring and its recent

acquisition of the Indian service firm Daksh (with its approximately 6,000 employees).

Hewlett Packard has built its global non-IT services to over 4,000 persons in the last three

years, largely through in-house hiring.

Software services is in most respects a headcount and labor-cost business; these

companies grow their revenues by hiring more persons. The multinational software services

firms have been experiencing increasing pressure on costs due to competition from

developing nation producers, particularly from the Indian service giants such as Infosys,

TCS, and Wipro. This has forced the multinationals themselves to secure lower-cost offshore

labor. Service firms such as Accenture, ACS, EDS, IBM, and Siemens Business Services

operate globally, but only in the last five years have they found it necessary to have major

operations in developing nations to decrease their labor costs. Today, the larger service

firms, including Accenture and IBM, are rapidly increasing their headcount in a number of

developing nations, particularly India. At the same time, these firms are holding steady on

their developed nation headcount or gradually drawing it down. Given the ferocious

competition in software services, there is little possibility that prices will increase

substantially. This suggests that, for the large multinationals, the offshoring of services will

continue to increase in both absolute numbers and percentages of their global workforce.

Next we consider firms headquartered in developed nations that have software operations

but are not part of the software industry sector. This is the enormous and eclectic group of

companies that provide all the non-IT goods and services in the economy. Software is now

at the heart of value creation in nearly every firm, from financial firms such as Citibank, to

manufacturing firms such as General Motors. Customizing, maintaining, and updating IT

systems has become an increasingly significant expenditure for businesses in developed

countries, and thus firms are actively trying to lower these cost. One way to lower them is

to offshore the work to nations with lower labor costs.

It is difficult to estimate the amount of software work that is offshored by these

companies. Businesses often do not break out this particular kind of expense, and if work is

transferred to an overseas subsidiary, this is considered an internal transfer and may not be

reported at all. However, it is clear who does the work. If it is not an overseas subsidiary of

the company, then it is likely to be one of two other kinds of firms that provides the service:

a large service firm from a developed nation (e.g., Accenture, CapGemini, IBM, and

Siemens Business Services) or a firm from a developing nation (e.g. Infosys or TCS in India,

Luxoft in Russia, or Softech in Mexico).

It is not certain whether offshoring will lead to a decline in the number of software service

employees employed in the developed nations. In the current economic recovery, existing

Page 29

firm headcount throughout the IT sector in the United States appears to be stagnant. In

other sectors, limited data are available. For example, in financial services it is unknown as

to whether the increasing headcount in developing nations has had any impact on

employment in the developed nations. The most that can be said is that non-IT firms are

increasing their IT employment in developing nations to serve the global market, and this

trend is underway across many different firms, including industrial firms such as General

Electric and General Motors.

Next we consider software-intensive small firms, particularly startups, based in developed

nations. For small startups, offshoring is often a difficult decision, although more recently a

number of firms in the United States have been established with the express purpose of

leveraging lower cost offshore skilled engineers. For many smaller firms, an offshore facility

can be demanding on management time. This is especially true in India because hiring and

retaining highly skilled individuals is difficult. The protection of intellectual property, which is

typically the most important asset that a technology startup has, is problematic in India and

especially China. There is substantial anecdotal evidence that, despite these challenges,

under the pressure from their venture capital backers and the need to conserve funds, small

startups are establishing subsidiaries abroad, particularly in India, to lower the cost and

increase the speed of software development.

A pattern is emerging for US startups. They may initially use outsourcing to, say, an

Indian firm as a strategy, but many soon establish a subsidiary in place of the Indian firm.

They do this for a variety of reasons, including worries about intellectual property

protection, control of the labor force, and management efficiency. The minimum size of an

offshored operation is reportedly as few as 10 persons. If this report is accurate, then it

may be possible for many more small firms to establish subsidiaries in developing nations

than have done this so far. Unfortunately, data on the scale and scope of offshoring by

startups are unavailable.

It is tempting to view offshoring by startups (whether to an Indian firm, say, or to their

own overseas subsidiary) as an unmitigated loss of jobs for US workers. Nevertheless, the

real situation is more complicated. Lowering the cost of undertaking a startup could mean

that the barriers to entry are lowered, thus encouraging greater entrepreneurship. The jobs

created by this entrepreneurship should be counted against those lost by offshoring. So,

correctly estimating employment net effect of offshoring in the case of startups is very


Finally, we consider firms in developing nations providing software services to firms in the

developed nations. The availability of capable software programmers in developing nations

provided an opportunity for entrepreneurs and existing firms to offer programming services

on the global market. It was in India where this practice first began in a significant way.

Because telecommunications links were not so sophisticated, the Indian programmers

initially were placed in the US customer’s premises. This practice was profitable and

gradually expanded to include remote provision of services – often to do Y2K work—when

telecommunication improved and demand heated up in the late 1990s. These developments

created an environment within which major corporations were willing to experiment with

overseas vendors, and a sufficient number of these experiments were satisfactory. The

result was that offshore vendors, particularly Indian firms, were validated as candidates for

software-related projects. These projects also allowed offshore vendors, again particularly

Indian firms, to grow in headcount, experience, and financial resources, so that they could

undertake larger and more complicated projects.

Software services firms from a number of the developing nations have become players in

the global economy. The large Indian firms (HCL, Infosys, Satyam, TCS, and Wipro) are at

present the global leaders. However, in China, Mexico, and Russia there are growing

Page 30

software service firms that employ between 1,000 and 5,000 people. Currently, the firms

from other nations are not large enough to compete with either the multinationals

headquartered in developed nations or the large Indian firms. Medium-sized firms in other

geographies can, however, reduce the risk for customers of having all their offshore work

done in one country, where it might be interrupted by a natural disaster or by political or

military problems. The larger multinationals and Indian firms are also establishing facilities

in other geographies, particularly Eastern Europe and, more recently, Mexico.

Firms are leading a global restructuring of the geography of software and software

services production. They are experimenting with a variety of strategies meant to utilize

workers that have become available in the global economy. This is true of software product

firms as well as multinational and developing-nation software service providers. The impact

of firms outside the IT sector with large internal software operations transferring some of

the software operations to lower-cost environments has been less remarked upon; however,

should the current trend continue, this will have a substantial effect on IT employment.

These firms have already relocated a significant amount of work from high-cost to lowercost

environments, and this process appears likely to continue, and possibly accelerate, as

firms become more comfortable working in developing nations. The offshoring of startup

employment bears particular observation because the US high-technology economy in

particular is dependent upon the employment growth that small startups provide.

5. The Globalization of Research

IT research is concentrated in only a few countries. About a third of computer science

papers come from the United States alone. A few additional traditional centers of

concentration of IT research (Australia, Canada, France, Germany, Israel, Italy, the

Netherlands, Sweden, Switzerland, and the United Kingdom) account for about another


This is not surprising considering the large part of world Gross Domestic Product (GDP)

concentrated in these same countries. There is a correlation between Purchasing Power

Parity (PPP) Adjusted Gross Domestic Product and computer science publication. However,

the share of computer science paper production by scientists in the traditional centers of

concentration of IT research is more than 60% greater than their share of world PPP GDP

(65% vs 40%). In contrast, Brazil, China, India, Indonesia, Mexico, and Russia together

account for 27% of world PPP GDP, but only 7% of computer science paper production.

IT research was even more concentrated in the past than it is today. The initial bloom of

IT research occurred in a few select locations in the United States and a few other countries

in the aftermath of the Second World War. This concentration has been perpetuated by the

natural tendency of strength to build on strength. Particularly in the United States, this

bloom was driven by ample government funding and a significant migration of scientific

talent from the rest of the world. In fact, there is little doubt that government funding has

played an important role in most countries. For example, on a per capita basis government

funding is significantly larger in Sweden and Israel than in the United States. The pattern of

strength in only few countries is amplified by a general migration of scientists from

countries that do not support graduate education and research to countries that do.

Research-driven innovation is seen by many countries as a way to increase national

wealth and standard of living. Both developed and developing countries are attempting to

build up or shore up their research capabilities. This means greater competition among

nations in the research area, and in particular competition for talent. Until recently, the

United States had won the research talent competition, but that situation is changing. Due

to strong efforts to foster research on the part of a number of national and local

governments outside the traditional centers of research, IT research is slowly but steadily,

Page 31

and almost certainly inevitably, becoming more global. This globalization of IT research has

been accompanied by a significant increase in the production of PhDs outside the traditional

centers of concentration, and a reduction in the migration of researchers to these centers.

In the long run, there is no obvious reason why IT research should be any more

concentrated than world economic activity in general.

Globalization allows more and better people to participate in IT research. Increasing

educational opportunities around the world means that more people are able to realize their

research potential, thereby increasing the size of the IT researcher pool and the quality of

the best researchers. A freer worldwide market in research means that potential funding for

IT research can more easily be targeted to those that can most effectively and efficiently

create research results. Both of these trends increase the amount of scientific advance that

can be obtained from a given level of resources. There is little doubt that this is good for the

field of IT and for the world as a whole; however, while we gain as a group, localities and

individuals may end up suffering losses.

Globalization provides improved opportunities for people who live outside the traditional

centers of concentration of IT research. It also provides improved opportunities for the best

researchers, due to increased global competition for their services. It may, however, limit

opportunities for other researchers in the traditional centers of concentration, for whom

global competition may mean declining wages or even the loss of jobs.

6. Risks and Exposures

Businesses that make offshoring and outsourcing decisions increase their own exposures

to risk, and at the same time potentially create additional risks and exposures at many

other levels, all the way from individuals to nation-states. Many of these other communities

of interest have scant awareness that they are being exposed. For every risk of privacy

invasion into an employee database that an employer might fear, data about ordinary

citizens is exposed to tens of risks. Bank records, transaction records, call center traffic, and

service centers are all offshored today. Voluminous medical records are being transferred

offshore, read by clinicians elsewhere, stored and manipulated in foreign repositories, and

managed under much less restrictive laws about privacy and security than in most

developed countries. The higher exposure to terrorist incursion, sabotage, or extortion

attempts has not received wide discussion by companies employing offshore labor.

A basic principle of security is that the longer the supply chain and lines of

communication, the more opportunity there is to attack them. The inherent difficulties in

international data communications are compounded by jurisdictional issues regarding

regulation and legal responsibility. Offshoring risks include data communications

vulnerabilities, loss of control of business processes, loss of control over network

perimeters, increased network complexity, clashing security policies and procedures, gaps in

personnel security, and drastically diminished ability to know about and respond to security


What seems particularly lacking within many procuring companies is an overall line of

authority and responsibility for primary data records as they pass through one, two, or more

subsequent offshore companies that perform work on the data set or perform operational

tasks for one purpose or another. Such “hands-off” management responsibility cannot be

presumed to work in the best interests of anyone concerned with risk attenuation.

Risks turn into incidents through two basic kinds of action—accidents and intentional acts.

The vast majority of incidents that can be anticipated originate with threat actors: rogue

employees, hackers, criminals, organized crime syndicates, industrial espionage, unfriendly

Page 32

nation-states, and terrorists. Effective risk management strategies include security due

diligence, business due diligence, active risk management, and third party auditing.

Commercial risk from offshoring is multi-faceted and different from security risk. Business

issues are primarily operational—concerning productivity, efficiency, and quality. Business

managers everywhere struggle with costs, delivery times, and product quality. Geographic

and cultural spread can adversely affect the latter two even as costs seem to be reduced.

Communication paths become longer and more convoluted; they are more apt to suffer

distortion and error from language and cultural difference. Supply chain networks become

more diverse, less centralized, and hence less controlled. Protection from manufacturing

sabotage and theft becomes more difficult because of the breadth of the system.

Intellectual property protection becomes more porous as the infrastructure expands on an

international scale. Legal barriers and costs increase as companies cross international

boundaries, due to conflicting regulations, procedures, and practices. Safety issues are

exacerbated by decentralized operational logistics.

The most contentious and perhaps most challenging aspect of offshoring is its risk impact

on individuals. Individuals are often pawns in this global restructuring of business. They are

at risk of loss of privacy, loss of jobs, loss of property through identity theft and credit card

fraud, and loss of security. Moreover, they have little say in these business decisions and

little they can do to protect themselves.

Offshoring adds threats and vulnerabilities that do not exist in domestic outsourcing, and

increases vulnerabilities that exist in all inter-network commerce. Multiple legal jurisdictions

add new risks. Distance adds complexity and vulnerability because cyber-space is actually a

complex of real-world service providers in distinct jurisdictions with varying cultures, all

under cost pressures. A company acting under a business culture not easily known to clients

cannot be assumed to be exercising all the same precautions that might be common

practice in the client business’s country. As more and more countries provide offshore

services, the price pressures on providers of outsourced services increase. With increased

price pressures, the temptation to skimp on security measures gets stronger.

There are a number of steps that can be taken for protection. Data that is being

transmitted should be encrypted. Offshoring providers should be vetted carefully.

Companies should have security and data privacy plans and be certified to meet certain

standards. Service providers should not outsource work without the explicit approval of the

client. Mass export of databases should not be permitted. Data should be accessed one

record at a time and on a need-to-access basis. The database should be encrypted. Certain

types of data should not be allowed to be exported across national boundaries.

Offshoring can also place national security at risk by threatening both military and critical

infrastructure operations. For example, the United States and other countries’ IT-based

military systems have adopted COTS (Commercial Off-The-Shelf) product purchasing

strategies, shared national and international commercial infrastructures, and Internet

Protocol technologies to facilitate network-centric warfare systems. It is more difficult for

the buyer to gain insight into source and application code documentation for COTS products,

especially if the providing companies are offshore. Many COTS components and sometimes

whole systems are developed and maintained by providing companies, which may

themselves procure development and services from other nations with privacy, intellectual

property rights, security, diplomatic, and defense policies possibly at odds with the original

procuring country. Thus, a COTS strategy increases the possibility of a hostile nation or

non-government hostile agents (terrorist/criminal) being able to compromise the system or

services. Attacks can cause malfunction and destruction of critical infrastructure such as

transportation, power, and financial systems, and loss of citizen confidence in their

infrastructure and government.

Page 33

The offshoring of homeland security technology development and management systems

that send vital information such as biometrics, identification codes, tax and personal

information overseas are of critical concern. Until better controls of this information are

developed, this presents a risk to all nations. Further research in methods to secure this

data and the development of nation-to-nation and international treatment of both the data

and how compromises will be handled is vital.

Globalization is here to stay and so are its international effects. National security and

social effects can never be completely mitigated, but country-specific and international

strategies can be put in place. Problems cannot be solved until they are defined and

accepted as valid by a sovereign entity and its citizens. Topics needing national attention

include legislation, international agreements, policing, tariffs, Internet policies, and more

equitable tax-structure strategies for companies investing at home. Other topics needing

public attention include more formal government-commercial agreements and funded

research to address data protection and communications between stakeholders involved in

homeland defense and critical infrastructure.

7. Education in Light of Offshoring

Offshoring creates major changes in the demand for workers. Some countries need more

workers, others fewer. Offshoring also causes the set of skills and knowledge of workers to

change. Education is a tool that enables a country to provide the skilled workers that it

needs, and thus it can be the centerpiece of a national policy on offshoring. Developing

countries that are building up their software service export markets, such as India and

China, need to prepare growing numbers of people to work in this industry. The developed

countries are facing questions about how to revise their educational systems to prepare

their citizens for the jobs that will remain when other jobs have moved to lower-wage

countries. These developed countries also have to find ways of making their education

system serve to increase the technological innovation that has historically driven

productivity gains, new employment, and new wealth for nations.

The United States has a well-established and complex IT educational system. The

bachelor’s degree is the primary degree for people entering a computing career. While

degree programs appear under many names, five majors cover most of the programs:

computer science, computer engineering, software engineering, information systems, and

information technology. Although there are some differences among these five types of

programs, they are many similarities in providing foundational knowledge related to

computer programming, the possibilities and limitations of computers, how computers and

computing work in certain real world applications, various skills about communication and

teamwork, and other topics.

In addition to the five traditional kinds of departments, a variety of new academic units

related to computing and information technology have begun to emerge in US universities.

These include schools and colleges of computing that typically include the degree programs

in computer science as one component, new schools that are separate from computer

science and information science programs that fill an additional need in the computing and

information technology space, information schools that in almost all cases evolved from

library schools, and campus-wide multidisciplinary information technology institutes aimed

at fostering collaboration of faculty and students across departments. While they are not the

programs intended to produce ace programmers or deep technical experts, the mix of skills

and perspectives is a reasonable educational experiment to try to produce students well

suited for higher-value-added jobs. There is also rapid growth in degree programs offered

by for-profit universities, which provide a convenient entry to the profession for working


Page 34

Non-degree programs also play an important role in US IT education. They include

certificate programs, non-degree courses offered by traditional colleges and for-profit

organizations, training associated with specific technologies, and corporate training

programs. These alternative kinds of training programs appear to be growing rapidly, but it

is difficult to quantify their extent or growth. There are many different goals being sought

through enrollment in these non-traditional programs: training for a specific IT career,

career advancement within the IT field, move from a non-professional to professional IT job,

continuing education to keep technical skills current, or gaining specific product information

or usage skills. There is also training provided by corporate universities for employees,

customers, and suppliers, which might include technical training, background information

about the company or its industry, or core competencies such as learning skills,

communication and collaboration, creative thinking and problem solving, global leadership,

or career self-management.

Recent changes in Europe, under the Bologna Declaration, have the goal of unifying the

European educational system along the lines of American system of separate bachelor and

master degrees. The Bologna process provides a standardized sequencing of degree

programs, makes it less time consuming to obtain the first undergraduate degree, and

makes the system more open for students who received their baccalaureate degrees in

developing nations to enter masters programs without having to repeat some of their earlier

training. The Bologna initiative has stimulated new interdisciplinary and specialized studies

in computing within European universities, especially those incorporating domain-specific

knowledge such as bioinformatics and media-informatics, and has also created separate

programs in software engineering and telecommunications. The increasing uniformity of IT

education across Europe will provide additional incentive for offshoring work from higher to

lower wage countries within Europe; in the long run it may lead to a leveling of IT wages

across Europe.

The German model is particularly important since the German-speaking nations represent

approximately a quarter of the European population. There are some major voices in

Germany in opposition to the Bologna initiative. For example, the T9 initiative, by the nine

largest and leading technical universities in Germany, argues that the traditional model of

university education leading to a diploma after nine semesters has considerable advantages

over the system that leads to separate bachelor’s and master’s degree. It is unclear whether

this will lead to modifications in the Bologna model over time.

India, as the largest supplier of exported software services, faces a different set of

educational challenges from the United States or Europe, namely to ramp up its higher

education system to staff its rapidly expanding software industry. Soon after India achieved

its independence in 1947, a decision was made to invest a greater amount in higher

education than is typical for a developing nation, even though there was not enough money

to finance primary education for all. This decision was taken in part to support the efforts to

build an educated workforce for the heavy industry that India’s leaders envisioned would

provide an important part of its revenue base. The investment in higher education was

advantageous to India when it opened up its markets and began to participate more

extensively in global trade in the early 1990s. There have been many competing claims on

government funds, and the central government has not been able to keep up with the

increasing demand for higher education. Policies were liberalized in the early 1990s,

allowing the formation of new private institutions of higher learning, resulting in the rapid

development of private postsecondary education. Whereas only 15% of engineering seats in

university had been at private institutions in 1960, 86% are private today. The rapid

advancement of the private university system has created some problems. Quality varies

widely, from clearly substandard to the highest international quality, and the government

Page 35

has not established, much less enforced quality standards. Some Indians also object to the

high tuition and fees as being counter to the equal access goals of the nation.

Today the higher education system in India is extensive and rapidly expanding. It

currently includes more than 300 universities, 15,000 colleges, and 5,000 training

institutions. Nevertheless, only 6% of the college-age (18-23 year old) population is

enrolled in college or university. Some of the schools, such as the Indian Institutes of

Technology and the Indian Institutes of Management are world-class; but the quality falls

off rapidly after the top 15 schools. Total bachelor and master degree production in the

computing and electronics fields is approximately 75,000 per year. There are also some

350,000 students in other science and engineering fields at universities and polytechnics

receiving degrees each year, and many of them enter the IT industry upon graduation.

Training in the latest technology, English-language skills, and other work-oriented topics

are also important to the Indian software industry. This training is offered both by many

independent training organizations and some of the large IT companies such as Infosys and

Wipro, which run their own training operations.

China faces the same educational issue as India in building a trained workforce for its

software industry, but its approach is different, through centralized planning. When the

Communist Party came to power, it was committed ideologically to education and the use of

science and technology for economic development. Upon the establishment of the People's

Republic of China in 1949, the Western powers pursued a policy of isolating China; a byproduct

of this was China’s adoption of the Soviet Union’s model of comprehensive and

specialized universities and a large network of research institutes. In 1978, the Chinese

university model was reformed to one that more resembled that of the United States and

emphasized comprehensive universities. In the 1980s, China began sending many of its

brightest science and engineering students to the West, especially to the United States, for

graduate education. Nevertheless, the government research institutes within China are still

enormous and play an important role in graduate education. Until recently, only a very few

universities undertook research; their highest priority was pedagogy.

As in the case of India, Chinese universities graduate an enormous number of students

every year. In 2001, 567,000 students received their first degree, including 219,000 in

engineering and 120,000 in science. The quality of these graduates varies dramatically, but

the sheer volume means that China has a large reservoir of technically trained individuals.

Until 2001, Chinese universities neglected software studies as an academic discipline. At

the end of the 1990s, the Chinese government recognized that it had a shortage of trained

software personnel and called for improvement in Chinese software capabilities as part of its

central planning efforts. In response, 51 Chinese universities established masters degrees in

software engineering. These degree programs quickly attracted students. Including all the

different kinds of curricula, China is now training about 100,000 people per year for the

software industry. There are internal criticisms of the education, including overemphasis on

theoretical education, insufficient attention to practice, and lack of familiarity with

international standards.

There are many challenges to implementing an educational response to offshoring.

Consider the challenges in the United States. IT work encompasses many different

occupations, each with its own skill and knowledge requirements. There are five major types

of undergraduate degree programs in IT, and each would require revision in order to

address offshoring. There are similarly four different degree levels (associate, bachelors,

masters, and doctorate) to revise. Non-degree programs, such as certificate programs,

corporate training, and non-traditional universities all also play an important role in

preparing the IT workforce. There are multiple career paths in IT to take into consideration,

not just the traditional one from a college degree to a career in the same field. Universities

Page 36

are slow to make changes in their employees and their course offerings. It is hard for

national bodies to predict and match supply and demand for the IT workforce, so it is hard

for the higher education system to know how to set its production levels. The mission of a

university is not only to prepare tomorrow’s workers; there are other goals such as

research, preparing tomorrow’s teachers, giving students a liberal education, and teaching

them to think critically that must be considered when revising a university’s program to

address workforce needs. Offshoring itself is rapidly changing (from bodyshopping, to call

centers, to business process outsourcing, to knowledge process outsourcing and other

higher value added tasks), so how is a higher education system to know what occupations

to prepare its students for? These challenges mean that educational systems will have to

continually adapt to serve well their students and countries in the face of increased


Although the educational needs and issues may look different from different national or

individual perspectives, this study has identified six overarching principles that should apply

in developing as well as developed countries wishing to participate in the global software


There is a need to consider the levels of IT work that are predominant in the national or

multinational economy being served by the educational institution, and which are likely to

be predominant in the coming years. Software and IT work can be thought of as consisting

of a spectrum from the more routine (e.g. system and computer maintenance and support,

basic programming) through the more advanced (e.g. application programming that

requires knowledge of IT and specific applications, whether business, science, media or

otherwise, or sophisticated systems programming and IT architecture development) to the

advanced strategic (development of approaches that utilize IT to advance the organization

strategically and provide it with a competitive advantage). As computer science and IT

curricula are developed, particularly at the national level, it is important to consider the

levels of workforce preparation to which the curriculum is addressed. In nations that are

current recipients of offshored work consisting of programming and routine software testing

and maintenance, for example, it may be desirable to focus the curriculum more heavily on

the lower levels. This may change, however, as the roles played by IT professionals in these

countries evolve and the offshoring providers aim to perform higher level work. In countries

that are seeing their commodity IT work being offshored, it will be desirable for the

curriculum to prepare students for the middle and upper levels of IT work, where the ability

to merge computer science and IT with applications and strategy are important. This is

likely to lead to an increased emphasis on application knowledge and a reduced emphasis

on programming skills. It should be stressed that in all cases, however, the predominance

of a certain level of IT work in a certain nation or region is just a generalization; all levels

will exist in all countries, and students will be needed to move into all of these levels. It is

the distribution that will vary.

There is a need for CS education to evolve, whether due to globalization or not. The skills

and talents needed by software and IT professionals have evolved over the past half

century, independent of issues such as outsourcing and offshoring. In general, IT

professionals are more likely to work in an application-specific context than previously, and

conversely, less likely to work on computer-specific areas such as compiler or operating

system development. They are more likely to work on large software applications in teams

that include applications specialists, and depending on the organization, also to collaborate

with sales and marketing staff. They are also more likely to work in an environment where

they are expected to be masters of certain software platforms and interoperability

standards, and know how to reuse code. Thus in general, it will be increasingly important

that a computer science or IT education involves training that enables the student to work

on large-scale software applications, to understand important business, scientific, or other

Page 37

application areas, and be familiar with the tools and platforms that are increasingly the

standards in the international marketplace. It also is increasingly important that the

education emphasizes teamwork and communication skills, especially as they are practiced

in a geographically distributed fashion.

There is a need for education to begin to prepare students for a global economy and its

possible impacts on their careers. It is increasingly likely that an IT professional will be

working in a global context. This may include being part of a multinational team, or

collaborating with customers or suppliers from other parts of the world. Thus, it will be

increasingly important that an education in computer science and IT help prepare students

for this global workplace. Education that acquaints students with different languages and

cultures, whether through courses, study abroad, or other means, will be increasingly

beneficial. Finally, to the extent that English is the common language of the IT industry, the

ability of nations to educate their IT professionals to be fluent in English will be a major

factor in determining their success in the outsourcing economy and in multinational


Educational systems that help prepare students to be creative and innovative will create

advantages for those students and their countries. As the lower tiers of software and IT

work become more commoditized, creativity and innovation will become even more

important, particularly in countries that experience the loss of support and programming

work. The creation of new products and new businesses will continue to lead to the greatest

commercial and scientific successes, and even more, become the differentiator between

organizations and between nations. Historically, some educational systems are seen as

fostering creativity in students more successfully than others. One crucial differentiator in

fostering a creative mentality in students is the research component of the educational

system, and the participation of students at all educational levels in the university’s

research enterprise. Another differentiator is the degree of rote learning versus more open

problem solving. Nations that currently have an advanced research enterprise in their

university systems may increasingly see this as their greatest competitive advantage in

educating computer science and IT students for the higher tiers of the IT workforce. Nations

that do not include a research component in their university systems will need to consider

whether, strategically, the investment in developing this component and culture is needed

to attain their goals for the IT economies in their countries.

Educational systems that not only pay attention to current business and industry needs

but also provide a core foundational knowledge will create advantages for those students

and their countries. To cite two national examples, the Indian educational system has been

particularly good at teaching the latest technology that is needed in business and industry

today. The United States has been particularly good at teaching foundational knowledge

that is likely to serve a student through most of his or her career. Foundational skills help

students remain current, and not become obsolescent, as the technology changes rapidly

around them. Although the particulars of a new technology in the workplace may be

different from what a student was taught in school, a basic understanding of computing

principles and ways of addressing problems will remain current even as the particular

technologies change. Of course there needs to be a balance between fundamentals and

currently relevant technologies in the student’s education. In order to prepare students to

be productive workers when they enter the job market, it also is important that the

educational system pay attention to the current needs of business and industry and select

the technologies it exposes students to in order to address industry needs. This goal can be

achieved through respectful interchange between people in the academic and

industrial/business worlds. No IT education can possibly fulfill all of the student’s

educational needs for an IT career, however, and IT workers should expect to have to

Page 38

engage in life-long learning in order to keep up with the rapid pace of technological change

and the rapid changes in the way that organizations employ information technology.

A good educational system requires the right technology, a good curriculum, and good

teachers. Fortunately, personal computers are relatively inexpensive, software for them has

been commoditized, and fast, inexpensive broadband communication is readily available

most places in the world. Thus, the technology for training an IT workforce is within reach of

much of the world. The model curricula that have been designed by the professional

societies have been and should be used in many places around the world. There is probably

value in developing a process by which these curricula can have greater business and

industrial input and react more rapidly to changes in the way that IT gets used in the world.

Although adopted around the world, the model curricula have been designed primarily for

degree programs in the United States. If the professional societies truly aspire to be world

bodies and develop world curricula, they should pay attention to the needs of other

countries and their degree programs as well. The teacher problem may be the most difficult

one to address. For example, in the United States, there are serious problems with the

preparation of high school teachers who introduce students to IT, and several times in the

past (in the late 1970s and again during the dot-com boom of the late 1990s) American

universities had difficulty recruiting and retaining quality faculty because of the lure of

industrial IT positions, and had inadequate number of students obtaining doctorates, which

are required to become faculty members. In India, critics complain about the general quality

of IT faculty, salaries are low, and there have been no funds to enable research either by

the faculty members or their students. Inducements to improve the quality of the faculty

would be helpful in India, the United States, and other countries.

8. The Politics of Offshoring

Globalization, especially in its manifestation as offshoring, is a hugely disruptive force that

effects the national movement of wealth and jobs. In addition to the educational responses

to offshoring discussed above, countries might adopt political responses. Developed nations

might take political action to stem the loss of jobs and wealth to globalization, either

through protectionism or measures to make the country more competitive. Developing

nations might take political action to create an environment in which its software export

industry can flourish. Our initial focus here is on the United States, which is largest global

offshoring procurer.

Public policy debate about offshoring began in the United States as a result of the wide

news coverage of the report in November 2002 by Forrester Research that 3.3 million US

jobs would be lost by 2015 as a result of offshoring. The most common response to

offshoring in the United States has been actions by the executive and legislative branches of

the state and federal governments to create protectionist laws and executive decrees to

control the movement of work out of the country. Bills have been introduced that limit the

citizenship or visa status of workers allowed to do work for US organizations or require that

call center operatives working outside the United States inform callers of that fact. There

are reasons to question the legality and efficacy of this protectionist legislation. Some legal

scholars believe that most proposed state laws and executive orders will be ruled

unconstitutional because of the Commerce Clause of the Constitution, which leaves control

of international commerce agreements in the hands of the federal rather than the state

governments. Legal scholars also believe that proposed federal legislation on offshoring may

break existing international agreements. There is also a risk of retaliation by other countries

to protectionist American legislation.

A second policy approach has been to propose reforms to the H1-B and L-1 worker visa

programs. The purpose of these programs is to help US companies find skilled workers, but

Page 39

critics claim that they are being misused as part of a strategy that enables companies to

export jobs, especially to India.

A third approach is to ensure that US tax law provides no incentives to moving jobs

overseas. These proposals would normalize tax rules between the United States and other

countries so that US-based multinationals will have incentive to repatriate earnings to the

United States that they earn in other countries. Tax law is hard to enact; and even if it were

enacted, there would still be disparities because of costs of health care, safe workplace

legislation, and environmental protection.

A fourth approach has been directed at providing support to Americans who lose their

jobs through offshoring. In 1962, the US Congress passed the Trade Adjustment Assistance

Act to offer job training and extend the length of time of unemployment benefits to

American workers who have lost their job through trade agreements. There has been a

political and legal battle over whether the Trade Adjustment Assistance Act does or should

apply to software workers. Progressives want to go beyond this act and also require

companies to provide three months of notification to workers whose jobs are to be

eliminated because of trade, extend the term length of unemployment benefits, provide

wage insurance paid for by the companies that offshore work to make up some of the drop

in wages typical in the displaced worker’s next job, improve retraining and reemployment

services, offer temporary health care and mortgage assistance, and allow multi-year income

averaging on federal taxes.

A fifth approach is to improve the innovation base. The basic idea is that, although some

jobs will undoubtedly be lost to low-wage countries, America can produce a substantial

number of new jobs, including many of them that are high on the value chain, through

policies that create a climate of innovation. Innovation policy generally has four elements:

making it more attractive for foreign students and scientists to work in the United States,

improving the educational system in the United States, attracting US citizens to the science

and engineering disciplines, and increasing federal support for research and development.

There have been numerous criticisms that the United States is not now doing enough to

build that innovation base, and there are proposals under discussion by both Democrats and

Republicans in Congress, as well as suggestions from various non-profit organizations, to

create new innovation initiatives.

How do policy issues in other countries that offshore work compare to those in the United

States? Australia presents an interesting case study in the politics of offshoring in that

Australia offshores work but is itself a country that has benefited greatly from free trade,

both in terms of its important export markets for wheat, wool, coal, wine, education, and

tourism, and also for the range of products that are available to its citizens through imports.

Debates over free trade arose in Australia over offshoring in 2004. There was sharp

criticism from the opposition Labor Party to the lack of policies protecting Australian jobs

and workers. Interestingly, the Australian Computer Society published a policy paper that

advocated free trade and resisted any protectionist measures. Instead, it called for

improvements in existing government programs to help displaced workers with re-training

and re-tooling, check-lists that would educate Australian companies on the cost-benefit

analysis of offshoring so that they would not rush headlong into it, and changes in industrial

policy to enhance Australian R&D. The sitting Howard government was pleased with the

report and outlined its own policy initiatives, which included more government support for

displaced workers, an effort to increase foreign direct investment in Australia’s IT industry,

and various improvements in teacher training, educational programs, and educational


New Australian government data appeared this year, showing that many of the temporary

visas for skilled workers are held by Indians, and many of these visa holders are doing

Page 40

programming work. These numbers concerned the Australian Computer Society, and they

have taken harder-line positions on both the skilled temporary visa program (known as

“457” visas) and on a permanent residence visa program, known as the General Skilled

Migration Program. While still endorsing the basic immigration policy of the Australian

government, ACS has called for adjustments in the 457 system to make it fairer. It has also

called for the permanent immigration program (General Skilled Migration Program) to be

substantially reduced until the market can absorb ICT graduates from Australian

universities, Australian computer science enrolments begin to increase, and unemployment

levels for computer workers fall to the level of other professions in Australia.

Sweden provides an example of the policy stance of a Western European country that

engages in offshoring. The Swedish economy and welfare has benefited greatly from a long

tradition of free trade, starting in the late 19th century. The policy includes agreements

between employer and worker associations on the basic principles for wage setting and job

assurance and a commitment to overall Swedish industrial competitiveness in knowledgeintensive

and high-wage industries. This industrial policy caused Sweden to create one of

the biggest high-technological industries in the world; and it has among the highest rates of

investments in R&D and outputs in terms of scientific publications and patenting. Sweden

has also become one of the most internationalized economies in the world, having a high

dependence on foreign trade for its Gross Domestic Product. Part of its industrial

rationalization is through offshoring to countries with lower production costs.

On several occasions, specific industrial policy measures have been taken by the Swedish

government to support industries with low and decreasing international competitiveness. In

the 1970s, considerable industrial support was given to the steel, clothing, and marine

industries when they faced large-scale failures, but the measures turned out to be futile. As

a consequence, Swedish policy has to a large extent returned to the basic policy principles

of free trade, so in the current globalization trends Swedish policy is almost completely free

from protectionist and direct job-protection arguments. There have, however, been a

number of initiatives to improve Swedish competitiveness and counteract the negative

impact of offshoring. They are all related to a new national innovation strategy advanced in

the spring of 2004, which has three fundamental points: technological development and

R&D as the key to Swedish competitiveness, investments in large-scale public-private

partnerships to achieve centers of excellence in R&D for specifically targeted industries, and

reorganization and increased funding for R&D startups and growth of small and mediumsized

research-driven companies. Software is not explicitly mentioned in the plan. In

Sweden, software development and production is primarily embedded in other

manufacturing or service-providing value chains.

Turning now to the developing countries that export software service work, there have

been significant policy issues at the national and state levels that have shaped the climate

for the Indian offshoring industry. These include regulatory policy as it affects foreign direct

investment, taxation, building an infrastructure, protecting intellectual policy, data

protection and privacy, and education and training policy.

The regulatory history is the longest and most comprehensive of all Indian policies

affecting offshoring. From the 1950s to the early 1970s, Indian economic policy focused on

identifying ways for domestic companies to replace imports. Policies enacted in the 1970s

that severely limited foreign ownership in companies operating in India drove out some

multinationals, including IBM. Regulation in the 1980s promoted the development of the

hardware industry and identified software as a promising export business; however, India

had limited success in the 1970s and 1980s in building an indigenous IT industry. India was

forced to liberalize its economy in 1991 in the face of severe cash problems. The new

industrial policy included reduced licensing requirements in most industries, allowed foreign

companies to hold majority interest in Indian companies in many industries, provided for

Page 41

automatic approval for hiring foreign technicians and foreign testing of technologies

developed in India, and reduced restrictions on the ways in which mergers and acquisitions

could take place.

Tax policy also had a shaping effect on the Indian software industry. In 1981, the Indian

tax code was revised to establish tax-free zones on profits and gains for manufacturers,

including software manufacturing. In 1993, the law broadened the tax-free zones to include

various science and technology parks. The law was again broadened in 2005 to give tax

breaks to software firms outside these parks.

Infrastructure policy also shaped India’s software industry. Laws intended to build a

favorable infrastructure and reduce labor regulations and other bureaucracy for the software

industry were enacted primarily by individual state governments, mostly in the southern

part of India. The one infrastructure issue subject to federal governance was

telecommunications policy. Beginning in 1991, the telecommunications sector experienced a

series of deregulations that continued until recently. Deregulation enabled the Indian

software industry to have access to a completely modern telecommunications system with a

capacity and cost that enabled the offshoring service companies to be internationally


China provides an interesting contrast to India. China is a policy-driven society, and one

sees much more significant intervention of the state in the economic development of the

software industry in China than in India. The national software strategy in India has been

focused on the export service market, whereas the Chinese are interested in capturing their

domestic software product and service markets as well as participating in the export


Until the 1980s there were mainly local rather than national companies in China. Much of

the capital available to businesses was tied in one way or another to the state, and many of

the decisions on capital allocation were made at the local level. Since then, internal trade

barriers have been dropped, enabling companies to build scale and move into neighboring

markets. In recent years, the national government has promoted economic reform through

competition among provinces and growth for individual companies by access to capital

through the national stock market. Consolidation and focus on the international market has

not yet occurred in the Chinese software industry. As of 2002, there were over 6,000

software firms in China; only 19 of them had sales exceeding $120 million.

Chinese policy towards forming technological capabilities has changed over time. From

1978 to 1985, the focus was on central planning and state control. In the period from 1985

to 1991, the focus was on enhancing the innovation system through greater state support

for both public and private R&D. Since 1992, the focus has been on enabling marketoriented

reforms to improve the quality of research and the skills of the workforce, and to

broaden the focus on development beyond the defense and heavy technology industries.

The government has taken a strong hand is the development of trained personnel for the

software industry. This included not only new educational programs, as described above,

but also concentration of highly skilled software talent in certain geographic areas, by

having the government facilitate transfers of skilled software personnel to the chosen

places, including providing accommodation for their spouses and children. The Chinese

government has also provided incentives for overseas Chinese software workers, especially

managers, to return home through such incentives as cash payments, cars, houses, and


The Chinese government supports R&D in universities, research institutes, and to some

extent industry. The best known of these initiatives is the Ministry of Science and

Technology’s High Technology R&D Program, known more commonly as the “863 program”,

Page 42

which has provided more than a billion dollars of government funding for basic research

since 1986. Other programs to provide research support include the Development Fund on

Electronic Information Industry, an R&D Fund on Industrial Technology, and a Technological

Innovation Fund. Although the government has continued to support important state

research institutes, such as the institutes of the Chinese Academy of Sciences, there has

been an effort to make them less dependent on the state and encourage them to reach out

to obtain external funding sources.

The government has also taken steps to improve the competitive business environment.

China does not have a long history of controlling anti-competitive behavior in a

technological sphere, and it has thus had to pass a series of acts that protect a competitive

environment, making illegal certain kinds of behavior such as impugning another company’s

reputation, bribing, threatening, and dumping. There have been targeted tax reductions to

companies that meet certain sales and export figures. Exporting firms have been given

favorable terms on bank loans, export insurance, and taxes and duties.

China has one of the world’s worst software piracy problems. The Chinese government

has taken a series of steps to try to curb piracy. In addition to the general copyright law,

China has passed several laws targeted at fighting organized crime that is manufacturing

and distributed copies of pirated software. Government organizations are coordinating antipiracy

campaigns, and are being encouraged to be model citizens themselves by using no

pirated software. A registry system has been established, under which owners who register

their copyrighted software are given extra protections under the law. However, software

piracy remains a big issue.

Politics is one of the ways (together with education, consumer boycotts, and labor action)

that nations can respond to offshoring. The general movement has been to avoid

protectionist legislation. Australia and Sweden have completely espoused free trade even

though they risk some level of unemployment for their IT workers. In recent years, India

has moved away from its protectionist and isolationist politics of the 1960s and 1970s. The

United States has had a number of protectionist actions suggested, but most of these

efforts have not been enacted into law, and today there are calls for policies to enhance its

competitiveness rather than to protect its jobs by legal and economic barriers. China is the

most protectionist of the countries studied here.

All of these countries understand that they have to make their national laws conform to

some degree with global practices if they want to be players in the global marketplace. Thus

China, for example, has been willing to revalue its currency despite the short-term gain

from keeping it artificially low; India has eased many of its trade barriers; the United States

has entered into numerous international trade agreements; and Sweden has conformed to

international monetary policies.

All of the countries studied here recognize that there are certain risks of sending software

work across national boundaries. These include questions of intellectual property, privacy,

and data security. Europe has taken the lead in strong privacy policy, and India has seen

the economic value in meeting European and US standards on privacy. China is not so far

advanced in managing these risk issues as India is, but there is every reason to believe it

will have to do so if it wishes to continue to attract international business. China is

struggling with balancing openness of information with political control, and so far it leans in

the direction of control rather than individual rights.

For the developed countries that send work offshore, a common political approach is to

build new jobs and prosperity through policies that increase innovation. Sweden is

increasing government support for research and development, and there are calls for this to

be done in the United States. The two countries differ on parts of the innovation platform,

however. Sweden currently has an abundance of highly educated workers, so it is not

Page 43

interested in ramping up its educational system. The United States is facing declines in

foreign scientists studying and working there, as well as declining numbers of American

students studying technical disciplines; so an integral part of the innovation platform for the

United States is to improve the education system and attract foreign workers and students

(to the degree this is compatible with national security policies).

India and China have a number of similar policies for developing their offshoring

industries. Both are interested in ramping up their educational systems to supply an

adequate number of skilled workers for their IT companies. Both are concerned about

having adequate infrastructures (power, transportation systems, telecommunications) to

provide good service to their IT companies. Both have adopted a series of policies intended

to attract foreign investment. China has implemented policies to try to produce a reverse

Diaspora, so that native-born scientists who have been working primarily in the United

States and Europe return home to be part of the senior technical and business leadership in

their IT industries; India has achieved this same effect without explicit national policies.

India has more experience in developing policies to support the export software market

than China, but China is advancing rapidly and has a more centralized government-planning

model in place.

Page 44 W. Aspray and O. Berry took a leadership position in the writing of Chapter 1.

Authors: Alok Aggarwal, William Aspray*, Orna Berry*, Stefanie Ann Lenway, Valerie


Chapter 1: Offshoring: The Big Picture

Offshoring is nothing less than a revolution in the tradability of services.” (World

Investment Report 2004, p. 148)

1. 1 Introduction

In the United States today, there are two views about the offshoring of IT and IT-enabled

services. Some people, such as the television business commentator Lou Dobbs, see a

crisis in our midst. More than a million blue-collar manufacturing jobs in the United States

were lost in the last ten years, mainly to low-wage Asian nations. The solace in all this for

American policymakers had been that another kind of job – the high-paying, white-collar

jobs in the computer and other knowledge industries that had long been dominated by the

United States – seemed immune to competition from low-wage countries. But then the

pattern of job loss began to be repeated in the white-collar labor force as the software and

IT-enabled service sectors moved jobs to Malaysia, the Philippines, China, and especially

India. Dobbs and others called for protectionist measures to stop the hemorrhaging of

high-paying jobs from the US economy. They believed that offshoring was not only going to

do short-term harm to those who lost their jobs, but also long-term damage to the

individuals and communities losing these jobs.

Others disagreed, pointing out that when this work is sent offshore, although domestic

labor may lose in the short term, there are many winners in the high-wage country:

consumers through lower prices; companies through higher productivity, more competitive

pricing, and shorter time to market; shareholders through higher corporate earnings;

company executives through higher compensation packages; and perhaps a select group of

other employees whose jobs change to include more interesting tasks associated with

innovation and exclude much of the drudgery of mundane tasks. Many of the supporters of

offshoring believe that the individuals who lose jobs will be able to find other good jobs,

especially if they are given a safety net from the state consisting of temporary benefits and

retraining, and that the total number of jobs may actually increase over time through higher

productivity and greater competitiveness of the companies that send work to low-wage


This difference of opinion in the public debates over offshoring is also found among

professional economists. Economists are generally regarded as being in favor of free trade.

For example, one economist who has looked closely at the issue of offshoring is Catherine

Mann of the Institute for International Economics in Washington, DC. She argues that free

trade will eventually lead to greater prosperity for the nation. She points to the case of

computer hardware manufacturing where many manufacturing jobs shifted from the United

States to East Asia in the 1990s as having been highly beneficial to the American economy.

Western innovation, coupled with global sourcing, led to price reductions in products. This

led in turn to more IT investment in the Western nations, higher Western productivity

growth, and ultimately enhanced growth in gross domestic product. Mann believes the

offshoring of computer hardware manufacturing was one of the reasons for the robust

economy in the 1990s in the United States and argues that the long-term national economic

benefits from outsourcing software and services are likely to be even greater than the

benefits from outsourcing hardware manufacturing. On the other hand, both Paul

Samuelson of MIT and Ralph Gomory of the Sloan Foundation, working with William Baumol

Page 45

of Princeton University and NYU, have done analyses that show that high-wage countries

can lose through trade under certain circumstances. (Mann’s argument is discussed later in

this chapter, Gomory and Baumol’s in Chapter 2.)

Looking at this same issue from the perspective of a low-wage country such as India, you

can also see two perspectives. Offshoring work is the top growth area in the Indian

economy, and it is the driver of India’s international trade. Hundreds of thousands of new

jobs are being created, and even entry-level positions in this field pay much more than the

average wage. IT is seen as the way for India to leap from being a third-world economy in

the 20th century to a world leader in the 21st century. However, this IT workforce still

represents only a tiny fraction of the Indian population, and there is a backlash to all of this

change in a country with rich cultural traditions. The traditional family structure is

threatened as young people move to the high tech centers for work, have large disposal

incomes, and otherwise follow work practices that do not fit with traditional culture. This

economic growth has brought congestion, unbridled growth, and severe wage differentials

to cities such as Bangalore. The benefits of offshoring are unevenly distributed with little

benefit for the majority of the people in China or India who are rural, poor, uneducated, and

without English language skills. Some critics complain that government funds spent on

attracting and building the infrastructure for IT companies could be better spent on helping

poor and rural populations with clean drinking water, better primary education, and other

basic infrastructure.

Which of these pictures is correct? Is offshoring leading to long-term deterioration of

Western living standards or is it the means to greater productivity and prosperity in the

West? Is it the economic savior for low-wage countries such as India and China or is it the

death knell for another traditional way of life? This chapter will introduce the subject of

offshoring of software and services and provide a framework for understanding it and

related issues from the perspective of both high-wage and low-wage nations. The following

questions are addressed in this chapter; many of them will receive more detailed attention

in later chapters.

What do we mean by outsourcing, offshoring, and globalization of software?

How did offshoring come about?

How much work is offshored?

Which countries send work offshore and which countries do most of this work?

What types of work are sent off shore?

Why are firms interested in sending work offshore?

What are the technical, business, and other drivers and enablers of offshoring?

Why might a firm, a profession, or a nation not want to offshore work?

Is IT still a good career choice for people working in countries that ship IT jobs


1.2 What Do We Mean By Outsourcing, Offshoring, and Globalization

of Software?

It is important to be careful about the terminology used in this study. Outsourcing means

that a firm sends work to another organization to be done. Most outsourcing done by US

firms, for example, is work sent out to other US firms. The client company might have

parts made for them or have another company handle the cleaning of their office premises,

for example. Offshore refers to where the work is done. It is a term that applies best to the

Page 46

United States because, even though the United States does outsource work to Canada and

Mexico, most of its work is sent over the seas, largely to India, but also to China, Malaysia,

the Philippines, and many other places. Germany, for example, sends work across its

borders, especially to Eastern Europe, but there is no water – no shore – to cross.

Some of the work that is offshored is sent to entrepreneurial firms established in lowwage

countries. Thus a UK firm that sends work to an entrepreneurial firm in India, such as

Infosys or Wipro, would be sending the work outside their own company.

At other times, multinationals headquartered in high-wage countries operate subsidiaries

in the low-wage countries to work on products and services for their world market. The

multinational might do this by contracting for all the services offered by an entrepreneurial

firm located in the low-wage country, in which case the entrepreneurial firm is sometimes

said to be a captive of the multinational, and the multinational holds great power over the

entrepreneurial firm. The multinational might instead buy an entrepreneurial firm in a lowwage

country outright, or it might create its own subsidiary there. These subsidiary firms,

whatever their organization, represent an increasingly large share of the offshoring of

software services.

Multinationals sometimes open facilities in low-wage countries in order to better serve the

local market especially since the Indian and Chinese markets are expanding so rapidly, but

that situation is not the primary interest of this study. We are more interested in

multinationals that open operations in low-wage countries to serve the world market.

Offshoring is part of a larger trend toward the globalization of software under which

software products and software services are created throughout the world and sold

throughout the world. The aspect of globalization that involves moving work from highwage

to low-wage countries is the most important aspect of globalization for this report but,

over the coming years, other aspects of globalization are likely to become important to the

professional, business, and policy communities.

More precisely, we should differentiate between captive offshoring and outsourced

offshoring or offshore outsourcing as it is often called. For compactness of language, we will

often use the term offshoring in this report without consideration for whether the work is

done by a captive or entrepreneurial firm. Where it matters, we are careful about the


1.3 How Did Offshoring Come About?

In order to understand offshoring, it is worthwhile to place it in the historical context of

globalization and multinational corporations. The import of raw goods and agricultural

products from less developed nations and the export of manufactured goods by

industrialized nations goes back centuries to a time when transportation across long

distances became feasible. Over time, some countries placed tariffs and other protective

barriers on international trade to protect their markets or industries. The first period of

intensive globalization came in the nineteenth century when laissez faire economic theory

drove nations to reduce or remove tariffs that limited the movement of goods. Globalization

was also driven by the adoption of the gold standard by many countries in the second half

of the nineteenth century. Gold stabilized the value of money and greatly enhanced trade

across national borders. Globalization led to the concentration of industrialization in the

industrialized countries at the expense of their agricultural bases, specialization in the

manufactured products they exported, growth in population, and demand for greater import

of agricultural products from agriculturally oriented countries. Globalization led to a

substantial increase in wealth for the industrialized countries.

Page 47

This period of globalization ended with the onset of the First World War, and then an era

of protectionism ensued between the two world wars. The second wave of globalization,

which continues today, began near the end of the Second World War with a meeting in

Bretton Woods, New Hampshire in 1944 that led to the formation of The World Bank, The

International Monetary Fund, and the reestablishment of the gold standard. The World

Bank, whose original mission was the financial reconstruction of nations destroyed by the

Second World War, broadened its mission to include reducing poverty through the funding

of state governments to improve their educational, agricultural, and industrial systems. The

International Monetary Fund was formed to oversee the global financial system. This it has

achieved by making the international monetary system more stable and by helping out

countries with monetary problems by supplying them with financial and technical assistance.

The period since the Second World War has been characterized by a series of international

agreements to promote free trade. This period began with the General Agreement on

Tariffs and Trade (GATT). Twenty-three countries participated in GATT’s first round of talks

in Geneva in 1948, but by the Uruguay Round of talks in 1993, the number of countries

participating had increased to 123. The Uruguay Round of talks led to the formation of the

World Trade Organization (WTO) as a successor to GATT. Under the WTO, there have been

a number of different approaches to enhancing global free trade: reduction of tariffs, export

subsidies, and other trade restrictions; formation of free-trade zones; reductions of

restrictions on capital; and increased agreement among national intellectual property laws.

Country membership has grown from 26 in 1993 to 148 today. The net effect of all this is

to have many more countries participating in international trade and to provide conditions

that enable this trade to occur more easily.

Multinational companies, which are simply companies operating in multiple countries,

have played an important role in the globalization of trade. The first multinational was The

Dutch East Indies Company, formed in 1602. The rise of big business in the second half of

the nineteenth century, with its concomitant separation of ownership from management,

created many new multinational companies. Some of these nineteenth-century

multinationals were technology companies such as I.G. Farben, which started its chemical

business in Germany, and General Electric, which started its electric power business in the

United States. Within a few years of their founding, both of these companies were operating

in many different countries around the world.

The computer industry attracted firms from the business machines, electronics, and

defense industries but also included important entrepreneurial start-ups. A number of

companies from the computer industries became important multinationals. These include

General Electric (formed in 1895 and entered the computer industry in the 1950s), IBM

(consolidated in the tabulating business in 1911), Hewlett Packard (formed in 1939 as an

instrument maker and entered the computing industry in the 1960s), EDS (formed in 1962

to serve large users of computers), Microsoft (formed in 1975 to provide products in the

microcomputer software industry), and Dell (formed in 1984 to provide microcomputer

hardware). It is notable but not surprising that these companies all had their origins in the

United States. The United States has dominated the computer industry throughout its

history. In its hey-day, IBM alone held about 70% of the world market for mainframe

computers, for example. The United States also had the market lead in the electronics

industry (mainly because of its dominance of the radio and television industries and its later

need for components for the computer industry) and the semiconductor industry, which

grew as a spin-off from the invention of the transistor at the regulated US monopoly AT&T

and was closely coupled in its history with the computer hardware industry.

US dominance in the computer, electronics, and semiconductor industries continued into

the 1970s, but then some changes began to occur. Perhaps the most public story was the

emergence of Asia as a leader in the manufacture of electronics and semiconductor devices.

Page 48

In the 1970s and early 1980s, major US electronics products firms began to set up affiliates

in Hong Kong, Singapore, and Scotland to use high-quality workers (with wages lower than

US workers) to do labor-intensive assembly such as assembling circuit boards or assembling

price-sensitive products such as computer peripherals or telephones. At first, the

components were built in the United States and shipped to these assembly plants but over

time the assemblers began purchasing components from local sources. Eventually, their skill

levels increased and they began to provide turnkey services. One specific example is disk

drive manufacture which began to migrate from the United States to Asia in the 1980s;

today, very little of this manufacturing takes place in the United States.

A similar story occurred in the semiconductor industry. Beginning in the early 1970s,

American (and later European) semiconductor companies such as IBM, Philips, AT&T, and

Hewlett Packard began to move labor-intensive chip assembly to low-wage countries in East

Asia, including Singapore, Hong Kong, Malaysia, and Thailand. These chips were then

shipped back to the American or European electronics firms for assembly into final products.

During the 1970s, the American semiconductor firms kept semiconductor wafer fabrication,

circuit board assembly, and product-level assembly in the United States. But both

computer and electronics firms opened or expanded plants in Scotland and Wales to do

circuit board and product assembly for the European market in the 1980s. Scotland and

Wales were selected for their educated workers, an English-speaking workforce, and

government incentives to attract foreign direct investment. It also helped that wages were

lower there than in the United States. More recently formed American companies such as

Sun Microsystems, Silicon Graphics, and Cisco never vertically integrated their operations

but instead always used contract manufacturers such as Solectron and Celestrica and chip

fabricators such as Taiwan Semiconductor. These firms were located in the United States,

East Asia, and Scotland.

In the 1980s in East Asia, Singapore’s labor rates became too high and its companies

began to offshore the most labor-intensive work to Malaysia and Indonesia which had lower

wage rates. A similar phenomenon occurred in Hong Kong which offshored its laborintensive

work to China. Singapore and Hong Kong retained the work on circuit board

assembly that could be automated. They also began to add backward integration services

such as component and circuit design, circuit board layout and reconfiguration for better

manufacturing, and forward integration services such as testing, final product assembly,

packaging, shipping, and repair. With a few exceptions, the East Asian companies providing

these value-added services chose not to produce products that competed directly with their

American and European customers. By the end of the 1980s, East Asia had the capacity to

provide circuit boards and electronics products to the entire world. At the same time, the

United States retained and grew its business for higher-value, lower-volume electronics

products such as large computers and communications switching equipment. This work was

often done under contract to specialized contract manufacturers, such as SCI and Solectron,

that were housed in the United States rather than by the large brand-name electronics

product companies themselves.

As more and more of this manufacturing work was done in other countries, middle-class

jobs were lost in the United States. It is hard to count the exact number of manufacturing

jobs created outside the United States to serve the US market or the needs of US-based

multinationals, but the number is probably in the range of a million jobs over the past

decade. The labor force in the US Midwestern industrial states was especially hard hit.

While this caused a public outcry and led politicians to suggest protectionist actions as

mentioned earlier, some economists see a silver lining in these developments. For example,

Mann argues that a combination of technological innovation in the United States and the

increase of global sourcing and markets for hardware (IT, semiconductors, and electronic

components and products) led to price declines. These price declines led to greater

Page 49

investment in IT in the United States. This, in turn, caused increasing transformation of the

American workplace and an increase in the development of new products either

incorporating IT or using IT in its development or manufacture. These developments, she

calculates, caused half of the productivity growth in the United States during the 1990s and

translated into increased wealth for the United States on the order of $250 billion in the

period 1995 to 2000. Mann assumes that there can and will be a similar pattern of growth

for the software industry but that the scale might be even greater for software than


While there has been angst in America over the number of good middle-class

manufacturing jobs lost to Asia, there has also been a widespread belief that good jobs in

the software industry would always remain in the United States. However, in the late 1990s

and even more so in the past several years, there is a dawning recognition and fear that

these high-paying software and service jobs will be moved out of the United States as well.

Similar concerns are now beginning to be expressed in Western Europe.

A number of IT-enabled services are being offshored today. They range widely and

include, for example, reading X-ray images of patients, identifying risk for insurance

companies, and processing financial data, as well as testing, building, and maintaining

software for customers. Software was the first service sector to be offshored to a significant

degree. This is perhaps because it was easy to transport the work data and work products

using simple communications equipment (a telephone and a modem) and because there

was a significant wage difference for programmers between the United States (or Western

Europe) and India (or China). During the late 1990s, software offshoring seems often to

have been driven by labor shortages in the United States, especially associated with fixing

the Y2K problem and creating new Internet products and services during the dot-com boom.

When the dot-com bubble burst, offshoring continued – with cost as a major driver – and

began to represent jobs transferred overseas rather than jobs supplementing an insufficient

US labor market. The practice of offshoring became a political issue in the United States

only after the recovery from the 2001 recession was historically weak in its creation of jobs.

European concern about offshoring lagged behind US concern presumably because the

United States began to offshore first and has always offshored to a greater extent than


Firms have outsourced work for centuries, sometimes even to companies that are outside

their national borders. The first offshoring in the software and IT services sector began in

the early 1980s: US firms sent some credit card processing to the Caribbean and

established call centers there. Software centers provided software services to the PC

manufacturers in Malaysia at about the same time. However, there was no substantial

software offshoring industry until the 1990s. India, Singapore, Ireland, Israel, and Hungary

were all early entrants in the offshoring business. Despite some differences in focus from

country to country, described in a later section of this chapter, all of these countries

benefited from first-mover advantages. Every several years, as a new application area

became hot, the offshoring firms in these countries would turn their attention to this

application, moving from business downsizing/reengineering, to Enterprise Resource

Planning, to Y2K, to Euro conversion, and so on. These offshoring firms coupled this

strategy with an effort to move up the value chain through industry sector specialization in

order to deepen their expertise and build trusted relations with clients who would eventually

turn over progressively higher level and more profitable tasks for them to do.

The story of how offshoring began in the major vendor countries, such as India and

China, is told in Chapter 3. These case studies indicate that offshoring has meant several

different things. In India, for example, it began with body-shopping, the process of sending

trained programmers to work for a few months in another country on the client firm’s

premises. This was followed by a blended strategy in which some of the work was done on

Page 50

the client’s site and some at the vendor’s site in India. Then call centers opened. In the

past five years, facilities began to be established to carry out IT-enabled business processes

such as accounting. More recently, Indian firms have begun to move up the value chain to

do IT-enabled knowledge processing such as reading X-rays, conducting patent analyses,

and carrying out IT research and advanced development. The players in this story were at

first Indian entrepreneurial firms. But later, multinational firms came to play an important

role, sometimes through an Indian firm that did contract work for the multinational

company, but also through a firm purchased outright or started up by the multinational


The globalization of the marketplace is helping to drive offshoring. The Indian and Chinese

governments, for example, have taken many steps to ready themselves to participate in the

international software market. Software is seen as attractive to low-wage countries as a

way to bolster their economies more quickly than the boot-strapping strategies tried in the

past by developing nations. In fact, about one-quarter of all offshored shared-service

centers for European clients involve interactions with the development agencies of the

vendor’s country (World Investment Report 2004). These countries have used tax breaks,

marketing subsidies, grants, loans, reduced bureaucracy, and other techniques to attract

foreign business and foreign capital. China passed the United States in 2002 as the most

preferred location for foreign direct investment. Trade policy has been liberalized in these

countries, for example, by reducing or eliminating export taxes and licensing (see Chapter 8

for details). These governments have enacted policies to strengthen the public and private

education and training sectors (see Chapter 7). Subsidies have been provided for research

and development activities in their countries, especially for development work that is likely

to have a near-to-midterm payoff in new products or services. Governments are trying

harder to protect intellectual property which has been an especially serious concern to

Western businesses about China (see Chapter 6). The Indian central and state governments

have worked to improve basic infrastructures such as telecommunications, electric power,

transportation (both roads and airports), buildings and technology parks, and other

amenities such as international-class hotels, but the infrastructure started in a poor state

and the government is not efficient in these efforts. Table 1 describes the state of

infrastructure in Bangalore, India’s leading offshoring location.

Table 1-1: Bangalore’s Infrastructure for Conducting Offshoring Work

Electric power is unreliable, so most companies have backup generators.

Roads are congested and in ill repair (an hour to travel the 12 miles from center

city to the outskirts where the outsourcing companies have their campuses in Electronics

City and Whitefield).

Work has not yet begun on a new international airport.

There is a shortage of rooms in international quality hotels.

No mass transit exists (talking of elevated railway) so most companies hire their own

buses to bring employees back and forth to work.

Telecommunications infrastructure is improving rapidly (cell phones, satellite

transmission, transoceanic fiber optic cable).

Source: Fannin (2004)

Page 51

1.4 How Much Work Is Offshored?

The answer to this question is that nobody has very good data on the amount of

offshoring worldwide, whether one measures the number of jobs lost or created through

offshoring, the number or percentage of companies offshoring work, the number of

companies providing IT software services for export, or the monetary value of this work.

Chapter 2 provides a detailed analysis of the problems with the data. It also provides a

sample of the statistics about the extent and impact of offshoring in the United States,

Europe (with separate breakouts for the United Kingdom and Germany), and India. Adding

these numbers up gives some sense of the global situation. In Table 2, we provide a sample

of the worldwide statistics as already totaled up by others. One can see from the McKinsey

(2005) statistics that the actual number of jobs offshored is still a small fraction (less than

15%) of the number that could be offshored. Other statistics in the table make it clear that

one type of offshoring –business process services – is growing very rapidly, and that there

is room for considerable growth since only 30 percent of the largest 1000 corporations are

currently offshoring any work of this type. It is also clear that India is the major provider of

these services. We do not have good numbers for the amount of software service work

(software maintenance, testing, programming) being done independent of work for call

centers or business process outsourcing. Nor can we tell exactly how much of the

offshoring work is being done by independent firms and how much by subsidiaries of

multinationals, although it is clear that the latter are a large part of the total. The numbers

do, however, give a general sense of the scale of offshoring activity worldwide. Additional

information about the size of the Indian and Chinese shares of the world offshoring and

software markets is given in Chapter 3.

Table 1-2: The Extent of Offshoring Worldwide

Source Data Reported Statistic

McKinsey & Co. (2005) Amount of onshore

outsourcing worldwide as of


$227 billion

Amount of offshore

outsourcing worldwide as of


$10 billion

Amount of captive offshoring

worldwide as of 2001

$22 billion

Number of IT services jobs

globally that could be done

anywhere in the world as of


2.8 million

Number of service jobs

worldwide that could be

done anywhere in the world

160 million

Number of actual IT service

jobs in offshore operations

in low-wage countries as of



Page 52

Evalueserve (2004) IT offshore revenue

worldwide, April 2003-March


$17 billion (almost half from

India, almost one-quarter

from Ireland; includes IT

products shipped from


UN Conference on Trade and

Development (2004) (as

quoted on

Percentage of world’s

largest 1000 companies

offshoring business process



Value growth in offshore

business process

outsourcing worldwide


$1.3 billion in 2002 to $24

billion in 2007

Scholl (2003) (as quoted in

World Investment Report


Market value for offshoring

of IT services (not including

captive production for


$1.3 billion

McKinsey & Co. (2003) (as

quoted in World Investment

Report 2004)

Market value for offshoring

of IT services, including

captive production for


$32 billion

1.5 Which Countries Send Work Offshore and Which Countries Do

Most of This Work?

Countries that send software and IT-enabled service work offshore are primarily highwage

countries that have advanced service industries. These are also the countries that

have the largest amount of IT work. According to Datamonitor (May 2005), the global data

processing and outsourced market had a value of $246 billion in 2004 with North America

accounting for 43.6%, Europe 29.4%, Asia-Pacific for 17.8%, and the rest of the world

9.1%. The country that started the offshoring trend and that sends the most work offshore

is the United States. The United Kingdom, Germany, France, and other Western European

countries come next. Although Japan has an advanced economy, it does not offshore as

much work as the United States or the Western European countries. In an interesting turn

of events, Indian offshore companies have begun to open facilities in China (where wage

rates are lower than in India and a huge local market is opening) and Eastern Europe (to

take advantage of proximity to the Western European market –nearsourcing). The extent of

this phenomenon is limited and recent, and it is not clear whether it is a strategy for Indian

firms based primarily on obtaining more contracts or on taking advantage of lower-wage


Which countries do the offshoring work is a more interesting story. There are quite a few

countries that have tried to develop this business, and these countries vary considerably in

their skill sets, labor costs, cultural fit with the countries seeking to have work done, levels

of technical and business expertise, and type of work that they offshore. The four countries

that have the most established offshoring industries (accounting for 71% of the market in

2001) in order of market share are Ireland, India, Canada, and Israel (McKinsey & Co. 2003

as quoted in World Investment Report 2004). The public stories make one think that

offshoring work is all done in low-wage countries such as India and China. In fact, the

Page 53

majority of offshoring services have historically been provided by developed nations, and

Ireland still leads with a 25% share. However, as Arora and Gambardelli (2005) point out,

the value added in Indian offshoring is higher than in Ireland because so much of the Irish

work involves localizing US software products for the European market. Moreover, the

growth rate of the national software export industry is much higher in India than in Ireland

so the relative position is changing rapidly. Canada and Ireland do have lower wages than

the United States, perhaps 10 to 20 percent lower, but there are not the extreme wage

differentials as there are between the United States and India or China. So this is not yet a

north-south or developed/undeveloped nation issue although the trend is in that direction

(World Investment Report 2004).

An assessment by the consulting firm A.T. Kearney of the most desirable future locations

for offshore work placed India at the top of the list, followed by China, Malaysia, the Czech

Republic, and Singapore (A.T. Kearney 2004 as quoted in the World Investment Report

2004). The expected rapid growth in offshoring activity occurring in low-wage countries will

make the public perception of who does offshore work progressively more accurate. The

Kearney report listed Brazil as the leading offshore source in South America; South Africa in

Africa; Hungary, Poland and Romania in Central and Eastern Europe; and Canada and New

Zealand among developed nations. Ireland, Portugal, Spain, and the United Kingdom were

listed as the preferred destinations for offshore work within Western Europe.

Countries doing offshore work fall into four categories as shown in Table 3. First are

those countries that take advantage of their large capacity of highly trained/educated

workers and low-cost wage scale. One example is China which has established businesses

providing offshore work on embedded software and IT-enabled financial services. Another

example is Malaysia which is building up business at the lower end of the offshoring market

in call centers and IT-enabled back-office business processing services. The principal

example is India which is the fastest growing destination for offshore work and is involved in

almost every aspect of the industry from call centers to business process outsourcing, to

software maintenance and testing, to software research.

The second category consists of countries that have competitive advantage through their

language skills to serve a special part of the market. While it is useful in any kind of

offshoring work for vendor and client to be able to speak the same language, it is essential

that workers in call centers, for example, be able to speak fluently in the language of their

customers. Thus China, which has relatively few people who speak English fluently, is

unlikely to become a major provider of call centers to the United Kingdom or the United

States. The Philippines, Mexico, Costa Rica, Chile, and Morocco have taken advantage of

their bilingual skills in English and Spanish to open up call center businesses serving the

United States. South Africa is the leading offshoring nation in Africa because of its Englishlanguage

skills. Some countries from Francophone Africa (Mauritius, Morocco, Senegal,

Tunisia, and Madagascar) have recently started to provide call center and telemarketing

service to France. India, of course, has been able to build up its call center business in part

because of its English-language skills.

The third category consists of countries that take advantage of their geographic proximity

to a country that offshores work, so-called nearsourcing. The nearsourcing countries not

only are located nearby, making it easier for executives from the client firms to visit the

vendors, but there is often a shared language and culture as well. These countries

generally do not have extremely low wages, but their wages are typically lower than in the

country that is offshoring the work. Canada is a major nearsource destination for the

United States, providing many high-end services. Poland, the Czech Republic, Hungary,

and increasingly the Ukraine, Belarus, Romania, and Latvia are building nearsourcing

businesses to serve Western Europe, especially Germany. In a poll of 500 top European

companies in 2003, the German consulting firm Roland Berger found that 50% of European

Page 54

firms were planning to offshore to other parts of Europe and only 37% were planning to

offshore to Asia (Gumbel 2004). China is trying to establish a nearsource business for

Japan and Korea. (One could call the second category linguistic nearsourcing and this third

category geographical nearsourcing. Doing so suggests that there are other kinds of

affinities between nations that might make them want to do business with one another such

as a common heritage or legal system as exists between the United Kingdom and countries

in its former empire.)

The fourth category consists of countries that have special high-end skills. Like the

nearsourcing countries, the wage rates might not be as low as in India or China, but they

are lower than those in the United States or Western Europe. Israel provides offshoring in

the form of research and development for multinational corporations and niche software

products and services, especially in the security and anti-virus software markets. Ireland’s

offshore business is mainly in the area of packaged software and product development; it

hosts many multinationals who are building software products and providing IT services for

the European market. It also has a number of small Irish-owned companies operating

mainly in niche markets. China is beginning to develop high-end skills in the Linux

operating system, bioinformatics, and anti-virus software. Australia exports high-end, ITenabled

financial services. India is beginning to develop research and development

laboratories for various European and American-based multinational corporations. Also, one

should not neglect the United States which exports the highest amount of IT products and

services of any nation, mostly to Europe, and mostly in the form of packaged software and

consulting services.

Table 1-3: Nations that Do Offshoring Work

Strategy Principal Examples Others

Cost and Capacity China



Language Skills Philippines


Costa Rica


South Africa






Nearsourcing Canada


Czech Republic








Special High-End Skills Israel



United States




1.6 What Types of Work Are Sent Offshore?

Various kinds of work involving the use of information technology are being offshored.

Types that are of primary interest in this study include:

programming, software testing, and software maintenance,

Page 55

IT research and development, and

high-end jobs such as software architect, product designer, project manager, IT

consultant, and business strategist (the extent to which these jobs have been

offshored is an open question).

Because the focus of this study is on offshoring of software and services, we are not

primarily interested in the following kinds of IT-related work, even though they are

frequently offshored:

physical product manufacturing – semiconductors, computer components,


business process outsourcing/IT enabled services/knowledge process outsourcing

(e.g. insurance claims, medical billing, accounting, bookkeeping, medical

transcription, digitization of engineering drawings, desktop publishing, and highend

IT enabled services such as financial analysis for Wall Street and reading of

X-rays), and

call centers and telemarketing.

A detailed list of the various kinds of IT and IT-enabled services that are being offshored

can be found in the World Investment Report 2004 (p. 150). These include various types of

audiovisual and cultural services, business services, computer-related services, higher

education and training services, financial services, health services, Internet-related services,

professional services, and animation. Many of these fall outside the principal focus of this

study. In Table 4, we identify skill levels required for various kinds of IT and IT-enabled

services also taken from the World Investment Report 2004.

Table 1-4: Skills Categorization of Traded IT and IT-Enabled Services



Definition Examples Requires Comments

low Low entry

barriers in

terms of skills,



Data entry

Call centers

general formal



knowledge of

relevant language

basic computer


few economies of scale

little agglomeration

medium Complex

services that

require more

advanced skills

financial and






routine data



services such

as ticketing


training required

(perhaps in

training schools)

may offer economies of


may have agglomeration


Page 56

high Most creative

and skillintensive















Advanced skills at

high levels of


often with strong



stringent entry


involve agglomerated

economies with different

skills, enterprises, and

institutions interacting

with each other to share

work, stimulate

knowledge flows and

allow specialized skills to

be fully utilized

Based on Box IV.2 in World Investment Report 2004

Jobs that are at the greatest risk of being offshored are also often those most at risk of

being automated, in which case labor would be replaced by technology instead of by foreign

labor. For example, although it has not happened to an extensive degree yet, software

automation tools might help to automate low-end software development.

The situation is far from static. We described earlier how India first offered bodyshopping,

then software services, only later IT-enabled services, and most recently research

and development. There has been a similar change in the pattern of offshoring by firms in

high-wage countries. In the 1980s and 1990s, the typical pattern was for an IT manager to

hire an outsourcing firm to carry out some task that was not critical to the mission of the

client firm. It tended to be an application development that was highly structured, required

relatively little interaction and project management from the client, had clear deliverables,

well understood bidding procedures, and transparent risk to both the client and vendor.

Often the vendor was located near the client. More recently, the pattern has changed. The

outsource firm is hired not by the IT manager but by a higher-level executive such as the

CFO, CIO, or perhaps even the CEO. The task is more likely to be mission-critical to the

client. The applications are wide ranging, but they often include tasks that are less well

structured than in previous times; ones that require greater amounts of client contact and

project management and where deliverables, costs, and risk are less clear. The vendor is

as likely to be located in another country as nearby.

But what are the characteristics of work favorable to performance offshore? John Sargent

and Carol Ann Meares of the US Department of Commerce have provided an excellent and

detailed answer to this question that is adapted slightly in Table 5.

1.7 Why Are Firms Interested In Sending Work Offshore?

The public perception is that companies in the United States, Western Europe, and Japan

send work to India, China, and other low-wage countries principally because of the lower

labor cost. There is some truth to this perception. Companies want to maximize their

profits, and, in many cases, the lower cost of qualified labor in these countries is the

principal reason for making the offshoring decision. Sometimes companies begin offshoring

for cost reasons but continue for quality of work reasons. Sometimes something else drives

the initial decision to offshore, for example, the lack of enough qualified workers in the

Page 57

United States during the Y2K era. This section shows that the situation is complex. There

are at least nine reasons, low-cost labor among them, why companies send work offshore,

and often more than one reason is in effect in a company’s decision to offshore. Here is a

summary of those reasons. Chapter 4 gives examples of the ways in which particular

companies of various types use offshoring as a strategic tool.

Table 1-5: Characteristics of Work Favorable to Performance Offshore Through


high wage differential with similar occupation/level in destination country

high labor intensity

clearly defined requirements, little nuance

repetitive tasks

rule-based decision-making and problem solving

documented or easily transferred content and process knowledge

discrete, separable; low degree of interaction across different services, applications

low degree of personal interaction with end users, clients

stable applications with minimum of “firefighting”

long projected useful life to amortize offshore set-up costs

low-to-medium business criticality

less time-sensitive, longer transition periods

projects involving simple and standard hardware and software

digital, Internet-enabled

low setup barriers

low-to-medium technical complexity


projects in business areas in which offshoring is a broadly accepted concept

tightly defined work processes

stable process

Source: Sargent and Meares (2004). Note: as the Indian companies, for example, move up the value chain, the

characteristics of work subject to being offshored may change.

1. Reduced Costs and Increased Margins. In the modern, investor-driven, globalized

marketplace, there has been a compression of resources, both time and money, that

companies, new and old, have with which to make a new business model profitable.

One response to this compression has been to reduce costs. Labor costs are a major

portion of service and other knowledge-intensive businesses so it is natural to want to

reduce these costs. When a new software engineer costs $45,000 annually in the United

States and only $5,000 per year in India, even with many additional overhead costs

associated with offshoring, most firms anticipate substantial savings in sending work to

the Indian software engineer over doing the work in-house in the United States. In this

way, the companies can make their new business start-up funds last longer or increase

their profit margins. While the focus in the public perception is on the low salaries, costs

are also sometimes reduced because the offshore vendor has scale benefits in doing the

work. (Another response to this compression, to address the time issue, is given in point


2. Access to Skills. The United States has the strongest postsecondary system in the world.

It trains many highly qualified workers from both home and abroad, and it also imports

workers who are educated or trained in other countries. But the United States does not

Page 58

have a monopoly on highly talented, educated, and experienced workers. As China,

India, Russia, and Eastern Europe have joined the world market, there is now an excess

of educated workers in certain countries such as India, Ireland, Russia, and some

Eastern European countries at a time when the US math and science educational system

is slowing down its production. In 1999, for example, China graduated three times as

many engineers as the United States. In particular, we are seeing strong pools of talent

outside of the United States in the IT, telecommunications, engineering, and health care


Thus another reason for companies to send work offshore is the size and quality of the

available labor pool. The applicant pools available to the offshoring companies in the

leading offshoring countries in many cases have been larger and stronger than the

applicant pools available in the United States and Europe. In the late 1990s, many US

firms turned to Indian vendors because they had available programmers with the

knowledge of legacy systems to make Y2K fixes. Similarly, during the dot-com boom in

1999 and 2000, many US firms turned to offshore vendors to find enough people who

knew the Java programming language. There was an abundance of such people in

India, for example, not only because of the large labor pool but also because of the

tendency of the Indian higher education system to react quickly to the marketplace and

teach skills that are in current demand.

To take advantage of this labor pool, many of the best offshoring vendors spend

substantial money on the hiring process, going through a lengthy and rigorous screening

process to identify employees who have a higher average quality than those available

for the client firms to hire directly. NASSCOM, the Indian software and services trade

association, has expressed concern recently about the uneven quality of the Indian

educational system, contending that while there are still large numbers of graduates, not

all of them have the quality education that gives India this competitive edge in offering

this access to skills.

3. Experience. Companies from the United States and Western Europe sometimes choose

to send work offshore because other countries have greater experience in a particular

field than they do. This experience can be of four types:

A. Experience with a particular technology. For example, China already has the largest

number of mobile phones in the world and India may be the second in this regard by

2012; these countries have skipped a level by not putting the infrastructure in

landlines but investing more in the wireless domain. Hence, it makes good business

sense to do R&D on wireless in India, China, and other emerging wireless markets. A

similar situation pertains to Linux which is a part of the Chinese government’s

national technology policy. While India and China are not yet the world leaders in

these fields, they have a growing number of scientists and engineers with knowledge

of these fields, and the overall level of knowledge in the country is growing rapidly.

B. Experience with a particular scientific domain. There are, for example, several

countries that provide offshore services with strong labor pools in the biomedical


C. Experience with particular management issues. For example, several of these

countries have strong experience managing projects that operate multiple shifts per


D. Experience with cultural and marketing issues in emerging countries.

4. Time Shifting. Offshoring enables companies to offer multiple-shift services that may not

have been offered prior to offshoring. For example, US hospitals are using US-trained

Indian physicians to read X-rays in India in time to deliver the results to the US doctors

Page 59

the next working day. This move can increase patient service at a reasonable cost.

Offshoring medical services can also provide rural areas with access to affordable

medical services. Some IT companies have several offshore sites, located strategically

by time zone, that enable them to provide round-the-clock services such as help desks

and network monitoring, while requiring none of their workers to have to work the

graveyard shift.

5. Time To Market. Some companies offshore work in order to reduce the time to bring a

product to market. The types of work offshored for this reason include R&D, production,

and other parts of the supply chain. One reason that time to market can be reduced is

that companies can take advantage of time shifting. A design team in the United States

can work regular business hours and then turn the work over to their team in China,

which is just beginning its regular work day, to either continue the design work or do

code checking. Then the Chinese team can turn the work over to their Indian colleagues

for the next shift who work on it and turn it back to the US team to start the process all

over again. Another way to take advantage of offshoring to reduce the time to market is

to divide the work into self-contained tasks that can be worked on in parallel in several

locations. Yet another strategy for achieving faster time to market is to

compartmentalize the work into a set of tasks that require different skill sets and parcel

the work out to the teams around the world that would be most effective or productive

at doing a particular part of the work. With synchronization points, this modularized

work process can be used effectively to create one single larger product developed on a

distributed basis in a timely manner.

6. Market Access. Companies sometimes find it strategically attractive to have a market

presence in countries in which they would like to sell their products. As

Balasubramaniyan, general manager at Wipro Technologies, describes this issue:

“Offshoring also helps a company be closer to its global customers, thereby providing

appropriate offerings to its regional market and ensuring speedier problem resolution.

Developers and support personnel in the relevant geographies have a better

understanding of customers’ needs, regulatory compliances and regional preferences,

and can better implement the product or provide the service.” (Balasubramaniyan and

Guyer 2004).

7. Ability to Send Overflow Work. Many small IT companies, especially those in IT services,

are usually faced with “feast or famine” situations, that is, during any given period of

time, either they do not have enough work or they have too much work. These small

companies cannot afford to keep a very large workforce on their payrolls because they

cannot afford the payroll in lean times, and therefore they have to work with a minimum

workforce. However, this causes problems for the company when it lands a large project

that needs to be completed in a short period of time. These companies can benefit by

sending work to large offshore providers who can supply very capable professionals with

the right domain expertise at the right cost. Larger companies face this same problem.

Companies are unlikely to want to hire extra staff for a project that might only have a

six-month or one-year duration because of the cost of hiring and the morale problems of

having to lay these workers off at the end of the project. The use of offshore workers

enables a company to ramp up and down quickly without these problems.

8. Extending Venture Capital Money. After the dot-com and the telecom busts in 2001,

many startups, especially in the IT, telecom, and biotech areas, have found it difficult to

raise venture capital. Those that have been able to raise such funding – as well as those

who are working on a “shoestring” fund provided by family and friends – are left with

little choice but to make the funding stretch as far as possible. Lower-cost locations such

as Israel and India become very attractive for them, and so it is not surprising that by

March 31, 2005, more than 170 startups already had established their R&D centers in

Page 60

India. Often, the venture capital firms themselves are pressuring the companies to use

offshoring to keep costs down.

9. Other Business Reasons. Using offshore workers can have other business advantages.

Given the low cost of labor, a number of the better offshoring vendors have expanded

the ranks of their middle managers who have time to mentor and enhance the skills of

the lower level employees and identify and implement process improvements that make

the work effort more effective. The vendor might have access to tools that are not

available to the client either because they are proprietary or because they would be too

expensive for the client to buy but not too expensive for the vendor who can use them

for many different clients. Clients who are not in the IT business may have more time

to focus on their core business and maximize their overall profits if they offshore their IT

tasks. Some companies have found that because the offshore vendors are eager to

retain their business, there is a stronger focus on continuous business improvements

and customer service than if the work had been done in-house. Public sector

companies, who may be regulated against large cost overruns and have rigid work rules

that make hiring new employees difficult, may find that offshoring provides them with

new flexibilities.

1.8 What Are the Technical, Business, and Other Drivers and

Enablers of Offshoring?

Offshoring has been made possible by a collection of technological, business, work

process, policy, educational, and other changes over the past 15 years. The technological

changes are the ones that are most often mentioned in the discussions about the growth of

offshoring, but they are by no means the only ones.

(1) Telecommunications infrastructure. Since the late 1990s, there has been a dramatic

increase in the telecommunications infrastructure. As part of the dot-com boom, various

telecommunications carriers competed to increase satellite and optical fiber networks to the

point where there was a glut in the market after the dot-com boom ended and prices

plummeted. India now has readily available low-cost, high-bandwidth communication and

access to all the major telecommunications applications such as email, fax,

videoconferencing, and cell phone. Telecommunications capacity between India and the

United States grew from practically nothing in 1999 to 11,000 GBS in 2001. The cost of a

one-minute telephone call from India to the United States dropped more than 80% within

several years after January 2000.

(2) Changes in information technology. A number of changes in information technology

also changed the opportunities for offshoring. Low-cost computing power became readily

available. Software platforms became standardized: IBM and Oracle provided the standard

for database management, SAP for supply chain management, PeopleSoft for human

resource management, and Siebel for customer relations. Offshoring vendors could invest

in the purchase of a small number of standardized software platforms and train their

employees in their use rather than having to deal with possibly hundreds of proprietary

software systems. Workers could learn standardized skills that were then portable.

Training and skill certification became simplified. A similar effect was created by using

commoditized, inexpensive applications software packages. Standardization of data formats

and networking protocols made it easier to move large data sets from client to vendor.

Interoperability standards such as MDA, UML, CWM, CORBA, and OMA were established

during the 1990s, making it easier to modularize software.

(3) Pace of innovation. The technological changes mentioned in (1) and (2) can be

considered enablers. One study (Bartel et al. 2005) discusses technology as a driver of

Page 61

offshoring. It found that a high level of IT use in an industry is not a predictor of greater

outsourcing. However, an increase in the pace of information technology change does

increase outsourcing. The explanation is that firms are more willing to gain access to the

latest technology through their outsource vendor than by sinking fixed costs into a

technology that is likely to change with great rapidity. One would describe the dot-com era

as an era of rapid IT change, hence driving companies to outsource.

(4) The downsized corporation. Since the 1970s, businesses in the United States began

to move away from vertical integration of the corporation, shedding activities that were not

regarded as core competencies, through eras of reengineering and downsizing. During the

1980s and 1990s, more and more activities were pared from the list of core competencies

and subject to outsourcing. As IT systems became more standardized, they were seen less

as core activities. And as corporations focused more on core competencies, there was big

growth in outsourcing of functions outside the core.

(5) Other business drivers. There have also been some business drivers of offshoring.

When rival firms began offshoring, many companies felt that they had to offshore in order

to remain competitive. Companies looked for ways to cut expenses to deal with the

economic downturn that began in 2000. Venture capitalists began pushing startups to

incorporate offshoring into their business plan so that the burn rate on start-up funds was

lessened. Several high-profile business leaders, such as Jack Welch from General Electric

and Carly Fiorina from Hewlett Packard, became evangelists for offshoring. As experience

with offshoring mounted, some of the early mistakes were understood and some of the

early problems with bureaucracy and infrastructure were fixed. It became more acceptable

and less risky to offshore; offshoring was no longer restricted to the early adopters such as

Texas Instruments or General Electric. Business leaders began to recognize the value of

reengineering, both in cost savings and improved performance, that was often undertaken

when work was transferred from client to vendor.

(6) Intermediaries. The offshoring business created new specialty occupations and firms.

Some of them did part of their work on the client’s site and offshored the rest; some did all

the work offshore on their own premises. Some served as brokers, placing the client’s work

with one or more of a number of offshore firms. Others served as consultants, helping

companies to make the decision about whether to offshore, what work to offshore, and how

to make contact and close a deal with an appropriate offshoring vendor. These consulting

firms and brokers aggressively marketed the advantages – particularly the labor cost

advantages – of offshoring. The presence of these specialty firms made it possible for

smaller companies to offshore work by helping them with the management of offshored

projects in a cost-efficient manner. Another group of firms emerged to provide support

services to the offshore vendors: transportation services, catering services, access to

specialty knowledge workers, and the like.

(7) Changes to the work process. Changes to the work process have enabled offshoring.

Certain kinds of knowledge work have been digitized and business processes have been

reengineered, making them suitable for offshoring. The value chain has been divided into

separable work processes some of which can be routinized and made subject to offshoring

even when not all aspects of the process are amenable to offshoring. This kind of

fragmentation of labor process is much more easily done with software and services than

with manufacturing. There is also more personal acceptance of having old ways of doing

business transactions replaced by using the Internet or proprietary networks to acquire

services so there is less resistance to service at a distance. All of these reasons add up to

the fact that progressively larger amounts of work can be offshored each year.

(8) Higher education system. In the past, one of the great advantages of the United

States has been its higher education system. However, some of the developing countries

Page 62

are using higher education as an effective means to create a skilled workforce, and the

numbers are impressive in comparison to the United States. Model curricula, established by

professional organizations such as the ACM and the IEEE, have been adopted in many

different countries, and the computing machinery needed for classroom instruction is

inexpensive in these days of personal computing.

In India, for example, national policy since the Second World War has placed a surprising

amount of limited resources into developing the post-secondary educational system rather

than in supporting the primary or secondary systems. There are 160 universities and 500

institutes today in India offering computing degrees of one kind or another, and the number

is growing rapidly. This is not yet as large as the number of colleges and universities

awarding computing degrees in the United States where about 200 universities offer ITrelated

doctoral degrees and about 2000 colleges and universities offer four-year degrees.

However, the number of technical degrees awarded in India as opposed to the number of

degree-granting institutions tells another story. India is awarding a much higher

percentage of its degrees in technical fields than the United States is. Each year, India

awards approximately 290,000 engineering degrees which includes 120,000 information

technology degrees, while the United States awards a total of approximately 75,000

computing degrees at the bachelor’s, master’s, and doctoral levels. India has also rapidly

built up a set of institutions for training people for jobs in IT skills training and certification.

To cite just one example, the offshoring firm Wipro has established Wipro University with 70

full-time instructors. It trains 2500 workers per year in areas directly pertinent to Wipro’s

offshoring work.

The United States still holds some significant advantages over India in the higher

educational system. Although India has a much larger population than the United States,

only a small fraction of its population attends college (7% as of 1997). The Indian system

is strongest in IT skills training and undergraduate degrees. University research in India is

very modest, and India produces only 300 master’s degrees and 25 doctorates each year in

the computing disciplines, compared to 10,000 master’s degrees and 800 doctorates in the

United States. For many years, the United States has been considered the place of choice

for advanced degrees for people throughout the world, but this seems to be changing.

Because of visa tightening and attitudes towards the United States in the post-9/11 era, the

number of foreign students applying to graduate school in the United States has

plummeted. The United States also has decreasing numbers of domestic students studying

IT (or science and engineering subjects more generally).

(9) Free-market world economy. The development of a world economy since the Second

World War has provided the opportunity for creating a global software market.

International agreements such as GATT and increasing national participation in international

organizations such as the World Trade Organization, the collapse of communism, and the

liberalization of the economies of India and China have all contributed to making the

software market global and in making India and China major participants in this market.

(10) Immigration. Immigration has played a role in the growth of offshoring. A large

number of Indian and Chinese citizens came to the United States to study and many of

them stayed on to work. In concentrated high-tech regions of the United States, most

notably in Silicon Valley, communities of Indian high-tech entrepreneurs emerged and

bonded with other Indians in the high-tech community, and similarly for the Chinese. In

many cases, these technical entrepreneurs were the ones who started offshoring companies

or who were the go-betweens to ease the difficulties of doing business across so many miles

and such different cultures. US immigration policy, especially the H1-B and L-1 visa

programs, have enabled Indians and other foreigners to gain valuable experience and

contacts in the United States before returning to their home countries. Recently, there has

been a reverse diaspora of engineers, and China and India are both recruiting technical

Page 63

workers to return permanently to live and work in their native countries. China has been

providing all-expense-paid trips to China, holding job fairs in Silicon Valley, and recruiting

faculty members online to spend their sabbaticals working in China. India has been

providing salaries, benefits, and stock options that make living in India attractive to Indian

high-tech workers who had been working in the United States.

(11) Other factors. Other factors have also played a role in driving or enabling offshoring.

The fact that English is the language of education and business has helped make India more

attractive to US firms. That India’s accounting and legal systems are compatible with the

British and American systems has also been an attraction. An aging population in the

United States means that the country will need to reach outside its national boundaries for

its workforce of the future. The McKinsey Global Institute projects a US need for 16 million

additional workers from overseas over the coming decades in order to maintain the present

ratio of workers to retirees. Evalueserve predicts a 5.6M worker shortfall by 2010, including

a shortage of 970K IT workers (assuming no work is offshored and not counting

immigrants). This shortage would lead to higher prices, being less competitive, and loss in

Gross Domestic Product. US temporary visa policy (for H1-B and L-1 visas) has been

conducive to building the offshore vendor presence onsite in the United States and in

building networks of people between India and the United States especially in Silicon Valley.

For the offshoring of research, international projects such as the International Space Station

and Human Genome project have built ties, involved many countries in the international

research community, and made internationalization of the R&D process seem more familiar

and feasible. The following case (Table 6) of the Indian state of Andhra Pradesh and its

capitol city of Hyderabad illustrates actions taken by governments interested in building up

an offshoring industry.

Table 1-6: Government Action to Build an Offshoring Industry: Andhra Pradesh

In the 1990s, the Indian state of Andhra Pradesh and its capitol city of Hyderabad

developed a public policy to create an IT-enabled services offshoring industry in their

locale. They created a government agency with the double entendre acronym APFirst

(Agency for Promoting and Facilitating Investment in Remote Services and Technology) for

this purpose. The government provided free right-of-way land for laying fiber optic cable

and donated a 55,000 square-foot office building with reliable electric power service to

encourage IT- enabled services firms to locate there. In 1999, the government created a

new incentive policy that provided 25% discounts on power bills to IT firms, exempted

software from sales taxes, and provided a rebate on the cost of land by up to 20,000

rupees per job created. In 2001, the Indian School of Business was enticed to relocate to

the city. The state created the Indian Institute of Information Technology to provide IT

education and the Information Technology Enabled Service Training Institute to offer

courses in English and other subjects of value to the ITES offshoring firms. The state also

declared the ITES industry to be an essential service, thus prohibiting its workers from

labor actions (just as in other critical industries such as water and police).

Sources: Atkinson (2004), Dossani and Kenney (August 13, 2004), Balatchandirane (undated “…Hyderabad…”)

There are some reasons why companies might not want to offshore. These have been

grouped in Table 7 into eight categories: (1) the job process is not routinized sufficiently to

offshore the work; (2) the job cannot be done at a distance; (3) the infrastructure in the

vendor country is too weak for the work to be carried out there; (4) offshoring the work

negatively impacts the client firm’s workplace; (5) there are risks to privacy, security, and

intellectual property of the client company from offshoring the work; (6) there are not

workers in the offshore company with the requisite knowledge; (7) the cost of opening or

Page 64

maintaining the offshore operation is prohibitive; and (8) miscellaneous other reasons.

Although each specific reason is placed in only one category, a number of the reasons could

fit in more than one category. Note that there is a certain parallelism between work that

firms might offshore (Table 3) and work they are unlikely to offshore (Table 7).

Table 1-7: Reasons a Firm Might Not Offshore Work

Category Specific Reason

Job process is not


*Uncertainty about the nature of work; uncertain specifications in

some jobs.

*Project has a highly iterative development process.

*Applications involve complex processes that require frequent

intervention to fix algorithms or data.

*High-skill work such as research, process design, or business


*The work involves system analysis.

Job cannot be done at a


*Face-to-face interaction is required for the job.

*It is too difficult to coordinate the non-standardized parts of a

project if they are geographically distributed.

The infrastructure is too

weak in the vendor


*Telecommunications, transportation, or specialty vendors are

not adequate.

The offshoring impacts

negatively on the client

firm’s workplace.

*The company loses control of the work process.

*The company loses in-house expertise needed to maintain,

improve, or replace the offshored product or service.

*Worker morale in the client organization deteriorates because of

potential loss of job, loss of wage power, or deskilling of job.

There are risks to the

client company in

offshoring the work.

*The work requires security clearance.

*Giving the vendor’s employees VPN access to the clients

information systems makes security difficult.

*Data privacy and security are hard to control at the vendor site.

*It is difficult to ensure that the vendor will protect the client’s

intellectual property.

*The vendor may not be able or motivated to follow the privacy

and security mandates in legislation from other countries such as

HIPAA, Gramm-Leach-Bliley Act, Sarbanes-Oxley Act, California

State Bill 1386, and European Union Data Protection Directive.

*The vendor may not be able or motivated to meet professional

qualifications required to do certain kinds of work such as being

an accountant certified as required by the client’s country.

*Legal recourse to privacy, security, or intellectual property

problems is non-existent or unenforceable in the vendor country’s

legal system.

Page 65

There are not workers

in the offshore company

with the requisite


*Application domain knowledge is required to do the job.

*The work crosses multiple disciplinary boundaries.

*The work depends on craft or proprietary knowledge held only

by the client company’s staff.

*The work involves business as well as technical expertise.

Cost of opening or

maintaining the

offshore operation is


*The client needs to implement new bureaucratic structures such

as explicit authority relations, operating procedures, and incentive


*There is an extra cost for evaluating vendors, managing

contracts, improving security, travel, and severance pay for laid

off workers.

*Alternatives to offshoring such as downsizing, consolidation, and

reorganization are more cost-effective.

Other reasons *Cultural issues exist between the vendor and client countries

(social behavior, attitudes towards authority, language issues).

*Gain occurs from being located near to other companies doing

similar work (agglomerated economy), e.g., jobs in complex

functions that need to be located near one another to thrive,

adapt, and innovate such as in activities in corporate centers or

less routine consulting practices.

There are also reasons that professions or countries might want offshoring not to occur.

If low-level programming jobs are shipped overseas, then there might not be a viable career

ladder for IT workers to climb in order to attain the higher-end IT occupations that people

hope will remain in the high-wage countries. Salaries of IT workers in the client (highwage)

countries might be pushed down by offshoring. The ingredients for innovation

(including labor, capital, knowledge, facilities, and technology) are threatened at home since

innovation is widely regarded as the driver of higher productivity and standard of living for a

nation. The locus of entrepreneurship begins to move offshore.

1.9 Is IT Still a Good Career Choice for People Working in Countries

That Ship IT Jobs Overseas?

Almost every day one can find stories in the US press about people losing their IT jobs

because their positions were sent to a low-wage country. Many of these stories quote

talented young people who are choosing careers in other fields because they believe there

are no longer opportunities in IT. There are fears that it will not only be low-level

programming jobs that are sent to low-wage countries but also jobs that require higher skill

levels and are more highly compensated. If the world really is flat, as Thomas Friedman

proclaims, and a job can as easily be done in Bangalore or Beijing as in Boston, then even if

the job remains in Boston, eventually the wages will fall in order to remain competitive with

wages in other parts of the world. One study has shown that if you are one of those who

loses a job to trade, the chances are that you will be paid less in your next job (Kletzer


All of this sounds bleak, but consider some interesting statistics on jobs as shown in Table

8 and on salaries as shown in Table 9. They are both based on data from the US Bureau of

Labor Statistics, one of the most reliable sources available. There is some lag in collecting

Page 66

and analyzing data so the most recent data is only from May 2004. Note what David

Patterson, a computer scientist from Berkeley who is president of the ACM, has to say about

these numbers:

“Moreover, most of us believe things have gotten much better in the year since the

survey was completed. Does anyone besides me know that U.S. IT employment [in 2004]

was 17% higher than in 1999—5% higher than the bubble in 2000 and showing an 8%

growth in the most recent year—and that the compound annual growth rate of IT wages has

been about 4% since 1999 while inflation has been just 2% per year?” (Patterson 2005)

How could it be that, at the same time jobs are being shipped overseas, the number of IT

jobs in the United States is growing rapidly and is even higher than at the height of the dotcom

boom? There are several possible explanations, but we do not have adequate data to

identify the one at play. One possible explanation is that the very companies that are

sending jobs overseas are prospering from the lower costs of overseas labor which is

enabling them to grow and create new jobs in the United States and elsewhere. Another

possible explanation is unrelated to offshoring except that the background factors that make

it possible are the same background factors that make offshoring possible, namely, many

industries are being reorganized to make them more productive through the use of IT.

Catherine Mann, the economist from the Institute for International Economics mentioned

earlier in this chapter, has conducted a study of Bureau of Economic Analysis data for the

years 1989-2000. (More specifically, her data is taken from BEA Digital Economy 2002,

Table A.4.4) She has found a strong correlation for industry sectors between high

productivity growth and high investment in IT (Mann 2004). She has also identified a

number of sectors that still have low IT intensity and thus are poised to take off as IT is

introduced. These include health care, retail trade, construction, and certain services. As

IT becomes more pervasive in society, there are more jobs involving either pure IT skills or

combinations of IT skills and skills associated with a particular domain such as finance or

health care. Most of the forecasts suggest that perhaps 2 to 3% of US IT jobs will be lost

annually to offshoring on average over the next decade. With the expanded use of IT in

society, it is very possible that the total number of IT jobs will grow at more than a 3% rate

over the decade. Thus it is not surprising that the US Bureau of Labor Statistics forecasts

that three IT occupations will be among the ten fastest growing occupations over the

coming decade (BLS 2002).

Even if the IT job market is a growth area over the next decade, some types of jobs are

likely to fall off, probably including routine programming jobs. As explained in Section 1.8,

there are many reasons that companies do not send work offshore so there are likely to be

jobs in almost every IT occupation to be found somewhere in the United States; but

perhaps in some of these specific occupations there will be fewer jobs that there are today.

It is very unlikely that the United States will be completely devoid of even these most atrisk,

routine programming jobs ten years from now.

There are no fail-safe recipes for having a successful IT career, but there are many things

people can do to make themselves more attractive to employers. They can get a good

foundational education and keep up with current technology. They can improve soft skills

such as oral and written communication and teamwork skills. They can get management

training and experience. They can learn the processes of a domain in which IT is likely to

be increasingly important in the future such as in the health disciplines. They can be

prepared to work on tasks that are less routine and that require regular discretionary

judgment or that require regular interaction with others (e.g., with customers or domain

specialists within the company). They can seek out jobs that involve knowledge of trade

secrets or fundamental processes of the company or that are involved with national

defense. They can learn about other cultures, the technologies for doing work in a

geographically distributed fashion, and other things about managing distributed work so

Page 67

that they can take advantage of offshoring instead of being a victim of it. They can gain a

wide array of experiences so that they can be employed flexibly by a company and so that

they gain an overview of the way IT is being used in the company and its industry sector.

There are also some things that American (or British or German or Japanese) society can

do to assure that there continue to be good IT jobs for their workers. They can nourish the

innovation base that creates these jobs. This can be achieved by adequately funding

research and development, improving the educational system at all levels, making sure that

there continue to be opportunities for foreign scientists and technologists to study and work

in the country because of their important role in driving innovation, and developing and

enforcing rules for fair competition in the international marketplace. These issues are all

discussed in the policy chapter (Chapter 8).

Table 1-8: IT Employment in the United States (US Bureau of Labor Statistics)


May Nov. May Change, May 2003

to May 2004

Occupations 1999 2000 2001 2002 2003 2003 2004 # %

Computer and

Information Scientists,


26,280 25,800 25,620 24,410 23210 23,770 24,720 1,510 6.50%

Computer Programmers 528,600 530,730 501,550 457,320 431640 403,220 412,090 -19,550 -4.50%

Computer Software

Engineers, Applications 287,600 374,640 361,690 356,760 392140 410,580 425,890 33,750 8.60%

Computer Software

Engineers, Systems


209,030 264,610 261,520 255,040 285760 292,520 318,020 32,260 11.30%

Computer Support

Specialists 462,840 522,570 493,240 478,560 482990 480,520 488,540 5,550 1.10%

Computer Systems

Analysts 428,210 463,300 448,270 467,750 474780 485,720 489,130 14,350 3.00%

Database Administrators 101,460 108,000 104,250 102,090 100890 97,540 96,960 -3,930 -3.90%

Network and Computer

Systems Administrators

204,680 234,040 227,840 232,560 237980 244,610 259,320 21,340 9.00%

Network Systems and

Data Communications


98,330 119,220 126,060 133,460 148030 156,270 169,200 21,170 14.30%

Computer and

Information Systems


280,820 283,480 267,310 264,790 266020 257,860 267,390 1,370 0.50%

Computer Specialists, All

Other 130,420 130,420

TOTAL (The "Change"

columns do not

include "Computer

Specialists, All Other")

2,627,850 2,926,390 2,817,350 2,772,740 2,843,440 2,852,610 3,081,680 107,820 3.80%

Computer Hardware

Engineers 60,420 63,680 67,590 67,180 72,550 70,110 74,760 2,210 3.00%

TOTAL, including

Computer Hardware

Engineers ("Change"

columns do not

include residual

"Computer Specialists,

All Other")

2,688,270 2,990,070 2,884,940 2,839,920 2,915,990 2,922,720 3,156,440 110,030 3.80%

Page 68

Table 1- 9: IT Mean Annual Wages (source: US Bureau of Labor Statistics)

1999 2000 2001 2002 May-03 Nov-03 May-04

CAGR (1999-

May 2004)

May 2003 -

May 2004

Computer and Information

Scientists, Research

$67,180 $73,430 $76,970 $80,510 $84,530 $85,240 $88,020 5.60% 4.10%

Computer Programmers $54,960 $60,970 $62,890 $63,690 $64,510 $65,170 $65,910 3.70% 2.20%

Computer Software Engineers,

Applications $65,780 $70,300 $72,370 $73,800 $75,750 $76,260 $77,330 3.30% 2.10%

Computer Software Engineers,

Systems Software $66,230 $70,890 $74,490 $75,840 $78,400 $79,790 $82,160 4.40% 4.80%

Computer Support Specialists $39,410 $39,680 $41,920 $42,320 $42,640 $43,140 $43,620 2.10% 2.30%

Computer Systems Analysts $57,920 $61,210 $63,710 $64,890 $66,180 $67,040 $68,370 3.40% 3.30%

Database Administrators $52,550 $55,810 $58,420 $59,080 $61,440 $62,100 $63,460 3.80% 3.30%

Network and Computer Systems


$50,090 $53,690 $56,440 $57,620 $59,140 $60,100 $61,470 4.20% 3.90%

Network Systems and Data

Communications Analysts

$55,710 $57,890 $60,300 $61,390 $62,060 $62,220 $63,410 2.60% 2.20%

Computer and Information

Systems Managers $74,430 $80,250 $83,890 $90,440 $95,230 $95,960 $98,260 5.70% 3.20%

Computer Hardware Engineers $66,960 $70,100 $74,310 $76,150 $79,350 $82,040 $84,010 4.60% 5.90%

3.90% 3.40%

3.80% 3.30%

11 24,720 5.60% 4.10% 3.60% 3.00%

412,090 3.70% 2.20%

CAGR 2% 425,890 3.30% 2.10%

1999 $100.00 318,020 4.40% 4.80%

2000 $102.00 488,540 2.10% 2.30%

2001 $104.04 489,130 3.40% 3.30%

2002 $106.12 96,960 3.80% 3.30%

2003 $108.24 259,320 4.20% 3.90%

2004 $110.41 169,200 2.60% 2.20%

267,390 5.70% 3.20%

74,760 4.60% 5.90%

1.10 Bibliography

Aggarwal, A. 2004. Moving up the value chain, from BPO to KPO. Evalueserve (Oct.). ACM

Job Migration Task Force Meeting (Oct.) Chicago, IL.

Aggarwal, A. and Pandey, A. 2004. Offshoring of IT services –present and future.

Eavlueserve Business Research (July).

Agrawal, V., Farrell, D., and Remes, J.K. 2003. Offshoring and Beyond.

The McKinsey Quarterly, No. 4.

Aspray, W. 2004. (Offshore) Outsourcing overview. ACM Job Migration Task Force Meeting

(Oct.) Chicago, IL.

Atkinson, R. 2004. Understanding the Offshoring Challenge. Progressive Policy Institute


Page 69

Balasubramaniyan, M. and Guyer, L. 2004. Face Off: Do offshoring’s benefits outweigh its

drawbacks? Network World (July 5) 42.

Balatchandirane, G. Education and Training [source unclear].

Balatchandirane, G. Development of IT Industry in Bangalore and Hyderabad IT Clusters

and the Impact on the Local Economy [source unclear].

Bardhan, A.D., Jaffee, D., and Kroll, C. 2004. Globalization and a High-Tech Economy:

California, the United States and Beyond. Springer.

Bardhan, A.D. and Kroll, C. 2003. The New Wave of Outsourcing. Research Report 1103,

Fisher Center for Real Estate & Urban Economics, University of California, Berkeley.

Bartel, A., Lach, S., and Sicherman, N. 2005. Outsourcing and Technological Change.

National Bureau of Economic Research (Feb.).

Bivens, J. 2004. Will the New International Division of White-Collar Work Make the U.S.

Rich? ACM Job Migration Task Force Meeting (Dec.) Washington, DC.

Booz Allen Hamilton 2004. Outsourcing Globally: Trends and Implications of Offshoring for


Dossani, R. 2005. IT Services Offshoring to India: India’s Position in the Supply Chain. ACM

Job Migration Task Force Meeting (March) Palo Alto, CA.

Dossani, R. and Kenney, M. 2004. Thoughts for the ACM Working Group. ACM Job Migration

Task Force Meeting (Oct.) Chicago, IL.

Dossani, R. and Kenney, M. 2004. Offshoring: Determinants of the Location and Value of

Services. Sloan Workshop Series in Industry Studies (Aug.).

Dossani, R. and Kenney, M. 2005. Moving Service Offshore: A Case Study of an U.S. High-

Technology Firm. (Feb.).

Drezner, D.W. 2004. The Outsourcing Bogeyman. Foreign Affairs 83, 3 (May/June) 22-34.

Fallon, A.J. 1993. Foreign Outsourcing of the U.S. Electronics Industry. Executive Research

Project 521. The Industrial College of the Armed Forces. National Defense University. Fort

McNair, Washington, DC.

Fannin, R. 2004. India's Outsourcing Boom. Chief Executive (May) 28-32.

Friedman, T. 2005. The World is Flat. Farrar, Strauss, Giroux.

Hagel, J. III 2004. Offshoring Goes on the Offensive. The McKinsey Quarterly 2.

Haveman, J.D. and Schatz, H.J. 2004. Services Offshoring: Background and Implications

for California. Public Policy Institute of California (Aug.).

Heeks, R.B. 1999. Software Strategies in Developing Countries. Communications of the ACM

42 (June) 15-20.

Hira, R. 2003. Utilizing Immigration Regulations as a Competitive Advantage: An Additional

Explanation for India’s Success in Exporting Information Technology Services. Center for

Science, Policy, and Outcomes, Columbia University (March).

Hira, R. 2004. Offshoring of High Skilled Jobs: Emerging Global IT Business Model. ACM

Job Migration Task Force Meeting (Dec.) Washington, DC.

Huws, U., Dahlmann, S., and Flecker, J. 2004. Outsourcing of ICT and Related Services in

the EU. European Foundation for the Improvement of Living and Working Conditions.

Dublin, Ireland.

Page 70

Kletzer, L. 2001. Job Loss from Imports: Measuring the Cost. Institute for International

Economics Washington, DC.

Kletzer, L. 2004. Trade-Related Job Loss and Wage Insurance: A Synthetic Review. Review

of International Economics 12, 5 (Nov.).

Koehler, E. and Hagigh, S. 2004. Offshore Outsourcing and America’s Competitive Edge:

Losing out in the High Technology R&D and Service Sectors. Office of Senator Joseph I.

Lieberman (May).

Krishna, S., Sahay, S., and Walsham, G. 2004. Managing Cross-Cultural Issues in Global

Software Outsourcing. Communications of the ACM 47, 4 (April) 62-66.

Mann, C.L. 2003. Globalization of IT Services and White Collar Jobs: The Next Wave of

Productivity Growth. International Economics Policy Briefs, No. PB03-11. Institute for

International Economics (Dec.).

Mann, C. 2004. Global Sourcing and Factor Markets: The Information Technology Example.

ACM Job Migration Task Force Meeting (Dec.) Washington, DC.

Mann, C. 2004. What Global Sourcing Means for U.S. IT Workers and for the U.S. Economy.

Virtual Machines 2, 5 (July/Aug.).

Marcus, A. 2004. Insights on Outsourcing. Interactions (July/Aug.) 12–17.

McKinsey and Company. 2005. The Emerging Global Labor Market (June).

NASSCOM-Evaluserve. 2004. Information Security Environment in India. Business Research.

Patterson, D.A. 2005. Restoring the Popularity of Computer Science. President’s Letter.

Communications of the ACM (Sept.) 25-28.

Ramer, R. 2001. The Security Challenges of Offshore Development. SANS Institute.

Sargent, J.F. and Meares, C.A. 2004. Workforce Globalization in the U.S. IT Services &

Software Sector, Office of Technology Policy, U.S. Department of Commerce. ACM Job

Migration Task Force Meeting (Dec.) Washington, DC.

Slaughter, S. and Ang, S. 1996. Employment Outsourcing in Information Systems.

Communications of the ACM 39, 7 (July) 47-54.

Sturgeon, T.J. 1999. Network-Led Development and the Rise of Turn-key Production

Networks: Technological Change and the Outsourcing of Electronics Manufacturing. In G.

Gereffi, F. Palpacuer, and A. Parisotto, Eds. Global Production and Local Jobs. International

Institute for Labor Studies. Geneva, Switzerland.

United Nations Conference on Trade and Development. 2004. The Offshoring of Corporate

Service Functions: The Next Global Shift? World Investment Report (chapter 4). United

Nations, New York, NY and Geneva, Switzerland.

U.S. Bureau of Labor Statistics.2002. Fastest Growing Occupations, 2002-2012. Labor

Review Table 3 (Feb.)

Wagstyl, S. 2004. Budapest, the Next Bangalore? New EU Members Join the Outsourcing

Race. The Financial Times (Sept. 21 as reprinted in YaleGlobal Online.

Wessel, D. 2004. The Barbell Effect. Wall Street Journal (April 2) A1.

* Burt Barnow took a leadership position in the writing of Chapter 2. Page 71

Authors: Ashish Arora, William Aspray, Burt Barnow*, Vijay Gurbaxani

Chapter 2: The Economics of Offshoring

This chapter provides background material on the economics of offshoring. The first

section deals with the economic theory applicable to offshoring, including discussions of why

firms engage in offshoring and what economic theory indicates the consequences may be

for individuals, firms, and nations. The second section covers the extent of offshoring; it

first cautions the reader about the difficulties in tracking offshoring activities, and then

provides estimates of current, recent, and projected offshoring activity in the United States,

Europe, India, and other countries. The third section draws general conclusions about the

economics of offshoring and identifies data that it would be useful to collect in order to

better understand offshoring.

2.1 The Economics of Offshoring: Rationale and Potential Impacts

From a long-term view, offshoring is a response to two developments: technical change,

especially in IT itself, and international differences in population and economic growth.

Advances in IT have made it possible to trade in what were previously untradable (or

difficult to trade) services such as software development, support, and maintenance. In

particular, as discussed in Chapter 1, the development of low-cost and high bandwidth

communications links that connect most corners of the globe have facilitated a massive

increase in the potential to move information around the world with virtually no time lag

and at low prices. This has allowed service providers in countries such as India and the

Philippines and manufacturers in China to coordinate and communicate with their customers

instantaneously. This ability to communicate has made a whole new set of sourcing

opportunities feasible that were previously unattractive due to the high costs. More rapid

population growth combined with increases in education levels outside the developed

countries has meant that countries such as India, Brazil, China, and the Philippines have

large numbers of young and talented workers who face limited opportunities for productive

employment and therefore have received relatively low wages in the local economy. The

combination of these developments makes for fruitful opportunities for gainful trade,

sparked by the sustained growth in demand for IT talent since the 1990s in the United

States and other developed countries.

Several additional factors make it easier for some countries than others to provide

offshore services. Even relatively low-skilled service jobs generally require literacy, for

example, help desk workers need to be able to look up reference material when needed.

Thus, countries with more educated workers are more likely to be able to handle offshored

jobs. Speaking the same language as the client nation is also beneficial for jobs requiring

communication with the client country; thus, India has an advantage over China for

receiving offshored work from the United States.

Some of the confusion and disagreement about the extent of offshoring stems from a lack

of agreement about whether goods are included as well as services, lack of precision when

applying the terms offshoring and outsourcing, and disagreements about whether direct

foreign investments are included in offshoring. For a discussion of some of these issues,

see Bhagwati et al.(2004).

In a free market economy, offshoring decisions are made primarily by private firms

seeking to maximize profits. The decision to offshore the production of goods or services to

another country can be implemented in one of two ways. A firm may choose to source the

good or service from a foreign provider that could be either an unaffiliated firm or a captive

Page 72

organization such as a subsidiary. (A captive organization is one which has its operating

decisions dictated by another organization. This might be, for example, because the captive

organization is owned by the captor organization or because there are strong economic

incentives for them to follow the dictates of the captor.) A firm’s choice to produce a good

or service itself or to outsource it to an unaffiliated firm is often referred to as the “make

versus buy” decision and can apply both to domestic and offshore situations. It is worth

pointing out that the degree of outsourcing of IT services to specialists has been increasing

significantly in developed countries over the last decade, initially to domestic providers and

recently to both domestic and offshore providers.

Firms consider any offshoring that they undertake to be in the best interests of their

stockholders. The theory of comparative advantage indicates that, if each country

specializes in the production of goods where it has a comparative (relative) advantage, both

countries can enjoy greater total consumption and well being in aggregate by trading with

each other. In our context, if providers in countries such as India have a relative advantage

in the provision of software services, for instance, then it would be beneficial for US firms

that utilize these services in their operations to source the services from Indian providers

and focus their scarce resources on activities in which they have a relative advantage. Note

that we do not say whether these are captive or outsourced providers since that decision

will depend on a variety of factors such as whether the service creates valuable intellectual

property. Likewise, the gains from trade are generally beneficial to the service provider’s

host country. In countries such as India, employees at software firms have seen significant

increases in their incomes even as the number of employees such as software engineers

increases rapidly. Similarly, revenues and profits at these firms are growing rapidly.

Services can also flow from developed countries to developing countries. For example,

companies based in developed countries such as ABN Amro, ING Bank, Prudential, and

Citibank are capturing significant market share in developing countries in numerous service

industries such as banking and insurance, and displacing workers in less efficient domestic

companies in these countries. In these sectors, the know-how possessed by these firms

provides them with comparative advantage. As Chapter 4 discusses, some of the IT service

firms headquartered in developed countries such as IBM Global Services and Accenture, are

adding workers from developing countries in order to take advantage of low wages, talent,

and location enabling them to compete directly with Indian software service firms such as

Infosys, Wipro, and TCS in the global software services industry.

What the theory of comparative advantage does not mean is that all members of society

will benefit from trade. In general, imports of an input have economic effects that are

similar to those of an increase in the supply of the input, namely, lower returns to the

suppliers of the input, lower costs of production, and lower prices for consumers. If the

input were a service, the wages and salaries of those producing the service would fall, but

so would the costs for firms that are buyers of the service. In the exporting country, the

opposite effects hold. That is, the returns to the owners or suppliers of the service or input

increase and the wages of the employees at the service providers increase due to the higher

demand. However, there are costs as well. In the short run, assuming lead time to develop

and scale service capacity, providers will often transfer capacity from a domestic market to

service the export market, raising costs to the domestic consumers of these services.

There have been relatively few studies estimating the economic impact of outsourcing and

most of those that exist have been based on European data and focus more on the

outsourcing of intermediate goods rather than services. Several studies identify cost saving

as the primary motive for outsourcing of intermediate goods, for example, Egger et al.

(2003). For the United States, Feenstra and Hanson (2001) construct industry-by-industry

estimates of outsourcing (of intermediate products) between 1972 and 1992 and find that

Page 73

outsourcing contributed substantially to an increase in domestic demand for high-skilled,

non-production workers and their wages.

European studies, such as Gorg and Hanley (2004), have used plant-level data for the

electronics industry in Ireland for the period from 1990 to 1995 to show that offshoring of

services had a positive, though not highly robust, effect on productivity growth1. Likewise,

Girma and Gorg (2003) find a positive impact of outsourcing of industrial services1 on

productivity in the UK manufacturing industries during 1980-1992, although they are unable

to distinguish between international and domestic outsourcing.

The studies, however, did not consider offshoring of services. Recent work by Mann

(2003) provides a back-of-the-envelope estimate that the first wave of offshoring, which

focused on global sourcing of computer hardware, led to a reduction in IT hardware costs of

10 to 30 percent in the 1990s and an annual increase in productivity of 0.3 percentage

points during 1995-2002, or $230 billion in additional Gross Domestic Product (GDP). This

fall in IT costs would be reflected in higher profits for producers and lower prices for

consumers. IT production workers lose, while firms and consumers gain. She goes on to

argue that, since a larger share of IT costs accrue to labor-intensive tasks such as software

development and systems implementation, the productivity impacts of offshoring these

tasks can be expected to be significantly higher.

Economists believe that trade generally leads to significant gains to society. These gains

are not inconsistent with employment losses in specific sectors that will cause economic

pain to the workers affected. To achieve an equitable result, many analysts believe that it

is important to establish a safety net that provides income and training opportunities to

affected workers. (See, for example, Atkinson (2004); Bivens (2004); Kletzer (2004); and

Mann (2004).) Components of the safety net should include unemployment insurance,

wage insurance, and retraining. This topic is discussed in more detail in Chapter 8.

A key assumption underlying the theory of comparative advantage is that the economy

enjoys full employment. Thus, this theory is best thought of as a theory of the long-term

where workers displaced by imports or offshoring find work in other sectors. By contrast,

most popular discussions of the offshoring phenomenon tend to focus on questions such as

“where will the new jobs be created” and “can the workers be retrained for these new jobs”?

In general, peering into the crystal ball to predict where and what types of new jobs will be

created is both difficult and unrewarding. A dynamic economy, such as that of the United

States, creates and destroys millions of new jobs in response to changes in tastes, and

more importantly, innovations and advances in technology. The US economy creates and

destroys more than 30 million jobs each year. In 1999, 32.9 million jobs were lost and 35.5

1 The authors find that international outsourcing generally had a positive effect on productivity, of

which the effect on the level of productivity can be attributed to outsourcing of material inputs.

Similarly, for international outsourcing of materials inputs, Egger, Pfaffermayr, and Wolfmayr-

Schnitzer (2001) find outsourcing of material inputs by Austrian manufacturing firms to the Eastern

transition economies increases domestic growth in total factor productivity, more so in capitalintensive

industries than in labor-intensive ones. Egger and Egger (2003) find that a 1 percent

increase in outsourcing of intermediate inputs to the Eastern countries relative to gross production

induces a shift in relative employment by about 0.1 percent in favor of high-skilled labor. Egger and

Egger (2001) find that outsourcing of intermediate products by EU manufacturing firms reduces

productivity of low-skilled workers in the short-run and increases it in the long run, an effect which the

authors attribute to imperfections in the EU labor and goods markets.

1 This study defines “industrial” services as “activities such as processing of inputs which are then sent

back to the establishment for final assembly or sales, maintenance of production machinery,

engineering or drafting services, etc.” (p. 5). They do not include “non-industrial” services such as

accounting, consulting, cleaning, or transportation services.

Page 74

million new jobs were created for a gain of 2.6 million jobs. In 2003, there was a net loss

of 100,000 jobs even though 30.2 million new jobs were created (BLS Business Employment

Dynamics). There is no guarantee that the economy will continue to create these new jobs,

but we can take some comfort from the historical evidence that thus far it has managed to

do so. The key to job creation is of course the ability of the economy to rapidly generate

and adopt innovations, that is, new types of goods and services, and productivityenhancing

process improvements.

Innovation is indeed an engine of economic growth, and perhaps the most important

source of productivity growth in developed economies. When there is rapid technological

innovation, as in the case of IT, there is a significant spillover effect to users of the

technology when the price paid for the technology is lower than the value received. This can

result in significant productivity and economic growth in the user sectors as well. Given

that IT is a general-purpose technology, the users are many and varied.

In general, trade stimulates innovation and economic growth in both trading partners.

However, Gomory and Baumol (2000) and Samuelson (2004) argue that innovation

opportunities create new possible conflicts of interest between trading partners. For

instance, insofar as offshoring stimulates innovation and productivity growth in countries

such as India, and more likely, Brazil, China, and Israel, in goods and services for which

developed countries such as the United States or Germany enjoy a comparative advantage,

this will cause the terms of trade to become less favorable for the developed nation. Simply

put, the comparative advantage of the developed nation becomes less valuable over time.

As a result, offshoring may impose permanent losses in the developed nation. In other

words, even if free trade is the best policy, it may well be that free trade, by stimulating

innovation overseas, may impose long-term losses. However, Gomory and Baumol’s

analysis shows that this conflict of interest is present when the two trading partners are at

similar stages of development. Since most offshoring involves countries at very different

levels of development, this conflict of interest is presently unlikely.

In the IT services sector, there is a related concern in the developed nations, particularly

in the United States. Currently, it is efficient to offshore low-end IT services such as coding

or maintenance, with high-end activities such as requirements analysis, design, and

research and development remaining in the developed country. However, the concern is

that eventually the “labs will follow the mills”, and high-end IT activities will also move

offshore. Were this to happen, the developed country might cease to be the technology

leader. There is some anecdotal evidence that some process innovations are moving

offshore. For example, in laptops, it used to be the case that contract manufacturers made

product to the design specifications of US vendors. Today, many of these companies have

moved upstream to design the product. Intel is designing processors at its R&D facility in

India. Likewise, software services firms have moved upstream to provide increasingly

sophisticated software solutions from more traditional applications, and to business process

services such as accounts payable, human resources, and even medical applications. Thus,

in addition to the static, resource allocation efficiency from free trade, one must also look at

the impact on the fundamental capabilities that underlie innovation.

These concerns reflect possible scenarios, perhaps even plausible scenarios. However, it

is not clear how likely they are. There are offsetting forces as well. In parallel with

offshoring, the inflow of skilled and trained workers into the United States has grown.

Though perhaps these inflows substitute for native-born workers in the short run, in the

longer run, they create all-around benefits by raising innovation. Moreover, it could be

argued that even a loss of technical leadership in one area could be beneficial by allowing

scarce talent and resources to be allocated to more promising areas such as nanotechnology,

bioinformatics, or genomics. The post-9/11 trend of a reduced rate of

Page 75

immigration should be of concern to the United States given that many other developed

countries are seeing increases in immigration of qualified professionals.

Arora and Gambardella (2005) suggest that the scenarios that result in a loss of US

technological leadership are very unlikely. In many industries, the locus of production and

the locus of invention are physically separated. This is particularly true when the body of

knowledge underlying the invention process has a strong scientific basis. Building on earlier

work by Lamoreaux and Sokoloff (1996; 1997), Sutthiphisal (2003) studied the location of

production and invention in three different industries during the Second Industrial

Revolution, namely, textiles, shoes, and the electric industry. He finds that, in general, the

locus of invention did not shift with the locus of production as the latter moved to other

locations. Moreover, he found that the link between location of production and invention is

weaker in the more science-based electric industry. Using data from a century later,

Mariani (2001) studied the location of R&D and production facilities by the Japanese

multinationals in Europe. She found that in low- and medium-R&D industries, R&D labs are

more likely to be located close to production facilities than in more R&D-intensive industries.

Chapter 5 discusses some of the current patterns in the globalization of research.

Can the United States specialize and keep its comparative advantage in the higher end?

The starting point for this discussion is to note that there are two key resources required to

remain the center of innovation in software: access to talented designers, software

engineers, and programmers along with proximity to a number of large and technically

sophisticated users. The United States dominates on both counts. Recall that in the 1990s,

there was considerable concern about Japan’s software factories (e.g., Cusumano 1991),

but there has not been a single successful Japanese software product that has developed a

global market (if one excludes the software that is a key component of gaming devices

which is not sold separately). To the extent that students are misreading the tea leaves and

moving away from studying computer science, the United States in particular could face a

longer-term problem in having access to talented software professionals. This topic is

discussed in Chapters 7 and 8.

The size and the openness of its culture and economy have given the advantage to the

United States over Japan and Western Europe in attracting talent from around the world.

The United States has been and continues to be a large producer of IT human capital for the

world, especially at the graduate level, and some of the students who study remain to work

in the country after graduation. The country has also been a magnet for technical workers

trained elsewhere. The United States as a destination for study and work has abated

somewhat in the past several years partly because of the harsher governmental regulations

in support of national defense.

Another global advantage of the United States is that most lead users are US-based. New

software applications depend largely on knowledge about demand and about the

applications domain. This is especially true for the substantial fraction of software used in

running businesses and business processes. Proximity to business activities is crucial for

innovations in such areas. Indeed, the development of new commercial applications or

solutions is a very special comparative advantage of the United States. On a more limited

basis, this benefit also accrues to other developed countries such as Germany where the

software giant, SAP, is based and dominates the market for enterprise software. In general

though, US industry is the largest user of IT in the workplace.

Globalization may reinforce this lead because we find that innovative companies from

Israel, Ireland, and even India are likely to move their operations to the United States to be

closer to their users. Sometimes, venture capitalists push for such a move as well. Other

intermediating institutions, such as legal services and thick and well functioning labor

Page 76

markets, are also important sources of advantage enjoyed by the United States that are not

likely to be eroded soon.

There are counter-tendencies as well. In so far as these professionals (be they Indian,

Irish, Israeli or Chinese) have a preference for staying in their home country where the cost

of scientists and engineers is lower, the cost of R&D activities that are human-capitalintensive

and relatively less intensive in physical infrastructure will also be lower offshore.

More broadly, the United States has several distinct capabilities – the best universities

and research institutions, highly efficient capital markets, flexible labor markets, the largest

consumer market, business-friendly immigration laws, and a large and deep managerial

talent pool. As a result, the evolution of business in the United States has followed a

consistent pattern: launch innovative businesses here, grow the business, and as products

and services mature, migrate lower-value-added components and intermediate services

over time to lower-cost countries. As more countries and regions enter global trade with

highly skilled and capable labor pools, this increases the amount of competition that US

companies and workers face. This competition produces an increased premium on the

innovation and productivity of the US workforce.

To continue growing the wealth of the United States and its individual workers, labor

productivity must grow at a rate that equals or exceeds the growth of wages. Labor

productivity can increase in a variety of ways but generally occurs through an increase in

worker skills combined with increased innovation in products and processes. Essentially,

growth requires capital investment in technologies that increase productivity and a focus on

innovation that results in new or transformed existing marketable products and in more

efficient processes.

In summary, economists have argued on both sides of the offshoring and free trade issue.

The arguments are sophisticated and nuanced, and the results often depend on whether the

underlying assumptions hold in the current context. While a majority of economists are

proponents of free trade, the underlying question is an empirical one and can be answered

by analyzing reliable data. As someone once said, “The difference between theory and

practice is greater in practice than in theory.” Unfortunately, there is a lack of data to help

understand the phenomenon, and more importantly, there are no data collection processes

underway that would help in conducting a statistical analysis of the empirical evidence.

It is important to distinguish among effects at the country, firm, and worker levels.

Because it is voluntary, offshoring benefits the firms that undertake it. Workers, on the

other hand, sometimes lose substantially from offshoring because they cannot instantly

(and may never) transfer their skills to other jobs that pay as well. At the country level, the

benefits of trade often outweigh the costs, but we have seen that this is not always the


2.2 Data on the Current State of Offshoring and Projections for the


“…there are currently no reliable statistical indicators of the extent or nature of global

outsourcing.” (Huws et. al. 2004)

The report that triggered public concern about the impact that offshoring would have on

US lives was produced by Forrester Research in 2002, indicating that 3.3 million US service

jobs would be lost to offshoring by 2015. This report was followed by many additional

studies, each with its own numbers, produced by private consulting firms, federal agencies,

and economists from academia and think tanks. These numbers ranged quite widely, in

Page 77

some cases differing by a factor of ten. How are we to know which numbers are correct?

This section discusses some of the general issues concerning data about offshoring. It then

considers what the existing data tells us about the current state of offshoring and about

projections of future offshoring activity. The greatest emphasis is on the United States (the

leading sending country), then on India (the increasingly dominant recipient country), but

data are provided about all the countries actively involved in offshoring to the degree that

we could identify data. Worldwide data was given in Chapter 1 and is not repeated here.

What Are the General Issues We Should Consider in Evaluating the Accuracy of Data About


There are three basic points to consider in evaluating offshoring data. First, there is a

question of the definition of offshoring. Some reports include all service jobs, some include

a subset of the service jobs that pertain to professional and technical services (following a

category used by the Bureau of Labor Statistics (BLS) in the U.S. Department of Labor),

some include all IT jobs, some include only software jobs, some include IT-enabled service

jobs, some include other jobs with other criteria, and some reports are not precise in

defining what they are counting. Obviously, the numbers will vary considerably based upon

the definition used.

Second, it is not clear what should be considered a good measure of the extent of

offshoring. Many people count jobs or workers. In addition to the issue raised in the prior

paragraph about what kinds of jobs one has in mind, there are other considerations.

Suppose one wants to count the number of jobs lost in the United States or Western Europe

to offshoring. How does one know which jobs to count? Business decisions by companies

are complex and, while the press sometimes reports horror stories of employees being

asked to train IT workers from a low-wage country and afterwards being replaced by them,

it is rarely clear-cut whether a job has been lost directly because of offshoring. A company

might cut back on the number of workers in one location and add workers in another

location, or cut back on people in one occupation and replace them with workers in another

occupation. And this might be because of some good business reason other than labor

arbitrage that is taking advantage of the wage differentials in the two countries to save on

labor costs. It might be, for example, that one product line is declining and resources are

needed elsewhere in the company, or the company needs fewer workers in a particular field

because of automation of some aspects of the work or rationalization of the work process,

or because the company has a global strategy that it is trying to achieve and part of that

strategy involves building market presence in specific geographic locations.

One might instead want to count the number of jobs created in a low-wage country to do

offshoring for a high-wage country. There is, however, no necessary correlation, for

instance, between the number of Indian jobs created and American jobs lost. A company

might decide to hire more Indian workers to work on a project than the number of American

workers displaced because the cost of the Indian workers is so low and better results might

occur by dedicating a larger labor effort to it. An American worker and an Indian worker

might not have the same productivity rate because of educational level, work process,

infrastructure, technical tools available to support the work, or many other reasons. The

literature gives examples where American workers are clearly more productive than the

Indian workers, and other examples that show the opposite. In particular, work processes

often are reengineered before being implemented in India, and the reengineered process

sometimes leads to significant increases in productivity. Also, in some lower level jobs (for

example, working in a call center), Indian companies are on average able to recruit workers

to do these tasks who are much more highly educated than the average American holding

that job. There is also confusion in the statistics about whether to count only the jobs

Page 78

newly created that are focused on doing software work for export versus counting all new

and existing jobs with this focus.

One might want to focus on the monetary value of the business rather than on the

number of workers who perform the work. This, too, is difficult to measure. One of the

problems is that a significant portion of the work that is done for multinational corporations

is done by their subsidiaries in countries such as India. The work that is conducted is then

regarded as an internal operation of the multinational, and this may make it difficult to

identify the monetary value of the work performed offshore or even to know when to

consider something as offshoring and when to regard it simply as a product or service

developed by a multinational through several of its divisions located in different countries.

This is a serious issue in measuring offshoring because balance-of-payments data tells us

that intrafirm trade represents 71% of all business, technical, and professional services

imported into the United States and, moreover, in the period 1997-2002, it was increasing

faster than imports from companies not affiliated with a multinational (UNCTAD 2004).

Another issue is that offshoring of complex products or services often occurs by dividing

up the labor and having some of it done within the client company, some done by one or

more vendors, and some purchased as components. In these cases, it is often difficult to

value fairly the portion of the product or service that has been outsourced. A recent

European study addressed these issues:

In its legal sense “outsourcing” refers to a business activity, involving the production

of either goods or services, purchased by an organization from an external supplier

rather than internally. It is, in other words, “subcontracting.” However, in the

current context of rapid organizational change, determining what is “internal” and

“external” is increasingly difficult. Mergers, demergers, strategic alliances, publicprivate

partnerships, and a variety of different forms of organizational disaggregation

– including those resulting from business process re-engineering – are increasingly

common. If a company is restructured on the basis of separate cost or profit

centers, for instance, should transactions between them be regarded as ‘outsourcing’

or merely as internal accounting flows?” (Huws et. al. 2004, p. 3)

If one wants to focus on the long-term impact of offshoring, the appropriate metric might

be jobs lost or created, or the monetary value of offshoring business over time rather than

at any specific point in time such as today. This kind of data tends to be of two types. One

type analyzes the nature of work to determine the number of jobs that might be vulnerable

to offshoring without making any claims that all or any particular portion of these jobs

would be transferred from the high-wage to a low-wage country. The other type analyzes

the number of jobs that will actually be sent overseas (or the monetary value of actual

offshoring business) by a certain date. Even in cases where the methodology is sound and

soundly applied, projections about the future are much less likely to be accurate than data

about today’s or yesterday’s situation since it is difficult to predict all the factors that will

come into play over time.

These two types of analyses introduce additional data issues. Vulnerability analyses are

less prone to errors because they require fewer assumptions than the other types of

analyses. To conduct a vulnerability analysis, all that is required is to identify industries

whose work could be transferred offshore and count the number of workers and their

occupations in the identified industries. To some extent, the list of vulnerable industries is

subject to change as new technologies and price changes can affect the list of industries

that are vulnerable to offshoring. A more important problem with vulnerability analyses is

that they tend to produce very large numbers that may bear little relationship to the actual

amount of offshoring that will take place. For example, all manufacturing, mining, and

agricultural activities could be replaced by offshore activities. Thus, other than pointing out

Page 79

which industries and occupations are immune from offshoring, assessments of the number

of jobs vulnerable to offshoring provide a very high upper bound that may be of little

practical value. As described in the following, projections of how much offshoring will occur,

while more useful in theory, are plagued by additional problems.

In addition to issues with the definition of offshoring and what constitutes a good measure

of it, there is an issue concerning the source of the statistics. There are four major

categories of data providers: government agencies, trade associations, consulting firms,

and universities and think tanks. At least in the United States and Western Europe, the

national governments provide data that is among the most trustworthy. For example, the

US government collects large and often fairly complete data sets, taking advantage of its

ability to compel business organizations and individuals to report certain kinds of data under

penalty of law. Most of these federal agencies employ well-trained and experienced

economists and demographers who typically use appropriate methodologies and open their

methods and assumptions to scrutiny. In the United States and a number of other

countries, the data-collecting and reporting agencies have been relatively unbiased – not

subject, for example, to political whims but instead trying honestly to determine what the

data tells us. Some analysts have argued, however, that US government data and reports

may indeed be either inferior to data from other sources or biased. For example, the

Economic Policy Institute (2004) has noted that the Bureau of Economic Analysis (BEA) data

on US imports of software from India shows much lower levels and a different trend (flat or

declining compared to a rapid increase) than data provided by NASSCOM, the Indian trade

association (see Figure 1). Business Week Online (2005) points out how the tone in US

Department of Commerce reports on the effects of offshoring changed markedly when one

of their reports was updated with the authorship shifted from career staff to political


Figure 2-1. U.S. Software imports from India

Page 80

Projections, however, require more than simply collecting and analyzing data. Models

must be developed to extrapolate trends, and decisions must be made on turning points and

when new forces will affect trends. Thus, it is not surprising that projections related to

offshoring, and other labor market variables as well, are subject to error, particularly when

projecting over longer periods of time. Even federal agencies are often far off the mark with

projections especially if these projections extend far into the future. The BLS periodically

assesses the accuracy of its projections. The most recent BLS review of its general

occupational projections shows that although BLS does reasonably well in projecting overall

employment in broad occupational groupings, the projections do not do as well in dealing

with specific occupations (Alpert and Auyer 2003). In projecting employment growth from

1988 to 2000, the BLS projected an increase of 15.3 percent, but employment grew by 21.7

percent over the period. The most serious problems in projecting occupational employment

patterns resulted from problems in anticipating changes in staffing patterns. Examples of

particularly large errors include the category of gas station attendant, which was projected

to grow from 308,000 to 331,000, but actually fell to 140,000, and travel agents, whose

ranks were expected to grow by 54 percent but whose numbers declined by 6.2 percent

(Wessel 2004). An earlier outside review of the BLS occupational projections found that

BLS tended to underpredict the growth of occupations requiring a college education (Bishop

and Carter 1991).

Even when government agencies do a good job of collecting data and making projections,

they generally collect data that helps them assess issues that have occurred in the past so

when a new phenomenon arises (such as offshoring), it is not clear that the data that

federal agencies have been collecting and the analyses they have been performing will

answer the policy questions that now arise. This is generally true for both the federal data

from most countries and for the data collected by pan-national organizations such as the

United Nations, the International Monetary Fund, and the Organization for Economic

Cooperation and Development (OECD).

The data provided from the other sources is potentially more problematic. Trade

associations, such as the Information Technology Association of America (ITAA) and the

National Association of Software and Service Companies (NASSCOM), the trade association

for the software industry in India, have access to large data sets from their members.

However, the members of these organizations are not necessarily representative of the

totality of producers or consumers of information technology so the data from these

organizations may not accurately represent the full story of what is going on. Moreover,

these organizations are partisan to their members, and they may craft their data studies or

reporting of these studies in ways that are favorable to the interests of their membership.

The consulting firms have a strong interest in increasing their business based on their

statistics. It is in their interest to provide numbers that demonstrate to potential clients

that there is a problem that needs fixing. Most of the consulting firms keep proprietary the

method by which they produce their projections so it is hard to evaluate what their numbers

mean. In fact, many economists are skeptical of the methods these firms use. Much of the

alarmist data about the impact of offshoring on job loss in the United States and Europe has

been generated by the consulting firms. All of this discussion indicates that it is difficult to

get accurate data about the amount of offshoring currently going on and that is likely to

take place in the future.

What Does the Data Tell Us About the Size and Impact of Offshoring in the United States?

Although the Forrester Research numbers are the most widely quoted, it would be

preferable to be able to use other sources because Forrester is an interested party and the

firm is not forthcoming about its methodology. One would like to use US federal data, if

possible, for quality and objectivity. The most relevant federal data source is the Bureau of

Labor Statistics Mass Layoff Statistics (MLS) series. Unfortunately, the MLS has serious

Page 81

shortcomings for this purpose in that it both undercounts and overcounts layoffs. It counts

only relatively large layoffs (at least 50 employees within a five-week period) and only

layoffs at companies that employ at least 50 employees; thus it undercounts by not

counting all small businesses and by not counting companies that lay off people for

offshoring purposes in smaller numbers or over a longer period of time. The MLS also

overcounts in that its numbers include people who are laid off for many reasons, including a

number not relevant to our interest such as bankruptcy, financial difficulties of the

company, movement to other locations within the United States, or change of company

ownership. BLS reports that these latter reasons are more common than layoffs for

outsourcing, both domestic or across national boundaries.

For the IT industries (by which the BLS means the communications equipment,

communication services, computer hardware, and software and computer services), MLS

only identified 7,923 people affected by movement of their jobs during the period January

through September 2004. Of these job relocations in the first three quarters of 2004, 70%

of the jobs were moved within the United States and 80% were moved within the same

company. When the jobs were moved contractually to another company, 40% of the jobs

were moved outside the United States. These numbers appear unreasonably low.

A study by Bronfenbrenner and Luce (2004) used online media tracking and corporate

research to identify offshoring job losses in the United States. This study found that the

MLS grossly underreported job loss due to offshoring. It estimated that 206,000 jobs were

shifted overseas in 2002 and 406,000 were moved in 2004. The authors argue that their

method, while imperfect, probably undercounts job migration because not all losses are

reported in the media and their search tools did not fully capture job losses that are

reported only in the local media, a common place for such losses to be reported.

2Even if the MLS data does not provide exact numbers, it might serve as a representative

sample from which one could learn about trends (for example, the year in which losses peak

in a particular field of IT). If this sample is indeed representative, computer hardware,

software and computer services, and communications equipment had their peak losses in

2001, while communications services had its peak losses in 2002. The Bronfenbrenner and

Luce study discovered that the largest job losses came from the midwestern states in the

United States, and that there have been rapid increases in job movements in IT, call

centers, and white-collar jobs to India.

Another potential source of federal data on offshoring comes from the US Department of

Commerce’s Bureau of Economic Analysis (BEA). However, some economists believe that

the BEA numbers seriously underreport software imports to the United States. For

example, BEA reports the United States imported $76 million in software from India in

2002, whereas the Indian software trade industry association NASSCOM reports Indian

software exports to the United States at $2,800 million that same year (see Figure 1). A

small part of this difference can be attributed to differences in definitions but not nearly the

entire amount. The US Government Accountability Office (GAO) noted the shortcomings of

the BEA data on offshoring as follows: “In addition to the lack of quarterly survey data for

unaffiliated transactions and lack of quarterly product detail for affiliated services, there are

reliability issues related to the mandatory filing requirements and survey coverage.”

(Government Accountability Office 2003, p.62).

2 This study defines “industrial” services as “activities such as processing of inputs which are then sent

back to the establishment for final assembly or sales, maintenance of production machinery,

engineering or drafting services, etc.” (p. 5). They do not include “non-industrial” services such as

accounting, consulting, cleaning, or transportation services.

Page 82

The BEA data nevertheless show some trends. Imports of business, professional, and

technical services into the United States increased by 77% to $38 billion between 1997 and

2002. US investments in developing countries that offshore services were small compared

to investments in developed countries, and most services created abroad are sold to non-US


Turning to trade association data about offshoring and the United States, the Information

Technology Association of America (ITAA) is the principal provider. ITAA is an organization

that represents 350 US-based Internet, software, service, and telecommunications

companies. It reports that 372,000 software and services jobs were lost between 2000 and

2003, with only 104,000 lost to offshoring. ITAA estimated that 90,000 new software and

service jobs were created in the United States in 2004 due to increased economic activity.

The largest amount of data (that is also the most suspect data) comes from the

consulting firms. Table 1 provides information about the impact of offshoring on the United

States based on reports and projections from consultants. The numbers generally indicate

that 12 to 14 million IT and IT-enabled jobs in the United States are vulnerable to relocation

through offshoring. Annual losses range from under 200,000 to about 300,000 for service

jobs lost from the United States due to offshoring. The number of IT jobs lost is somewhat

lower than these estimates because the estimates include service jobs such as working in

call centers and possibly in other IT-enabled services such as business process and

knowledge process outsourcing. The numbers from the American Electronics Association

might seem to be contradictory to the other data, but it should be remembered these are

net losses in the industry so they include losses not only to offshoring but for other reasons

such as company downsizing or bankruptcy, and these losses are offset by newly created

jobs. The consensus seems to be that about 20% of US companies are currently offshoring

work but that the percentage will rise considerably over the coming years. Bednarzik

(2005) concludes that “employment trends by industry and occupation suggest that

offshoring in the information technology sector occurs, but not to a great extent.” These

numbers also do not take into consideration jobs that are created by offshoring.

The current value of offshore contracts from the United States seems to be in the $10 to

20 billion range, and most analysts believe there will be rapid growth in these numbers over

the coming few years. It should be remembered, however, that we do not know the

methods used to arrive at these numbers and how independent the data from one

consulting firm is from that of another. We do not know, however, of any body of extant

raw data that serves this analysis well.[BSB1]

Table 2-1: The Impact of Offshoring on the United States

Source Data reported Statistic

I. Current or Recent Offshoring

Forrester (2004) US service jobs lost in 2003 315,000

Forrester (2004) US service jobs lost by end of



Goldman Sachs (2004) US jobs lost in past three years 300,000 to 400,000

Page 83

Dossani (2005)[BSB2] Software workers in the United


1,200,000 software

engineers; 500,000


Evalueserve (2004) IT jobs offshored year ended

March 2004

212,000 (60% to India)

Evalueserve (2004) Call center jobs offshored year

ended March 2004

136,000 (90% to India)

Gartner (2004) Fortune 500 companies

expected to offshore some IT

work by end of 2004


Gartner (2003) IT industry and employment 500,000 jobs by end of


Meta Group (2004 Annual IT

Staffing and Compensation


US companies using offshore

labor in software


American Electronics

Association (2003

Cyberstates report) [Seeley


Jobs lost in 2002 in the US

software services sector

30,000 (compared to

146,000 the year before)

American Electronics

Association (2003)

Jobs lost in 2002 in the US

software industry


American Electronics

Association (2003)

Jobs gained in 2002 in US in

high-tech R&D


Washington Alliance of

Technology Workers

(, 2005)

Jobs lost in the US IT sector

March 2001 to April 2004


Washington Alliance of

Technology Workers

(, 2005)

Percentage of IT sector jobs in

San Francisco area lost March

2001 to April 2004


United Nations Conference on

Trade and Development

(World Investment Report


Average percentage annual

growth in US imports of

computer and data processing

services, 1992-2002


Bajpai et al. (2004) Percentage of companies that

have offshored work (survey is

mostly but not exclusively of

US companies)


Bajpai et al. (2004) Percentage of companies that

have already or plan to

offshore work


IDC Value of offshore contracts

from US in 2005

$17.6 billion

Page 84

ITAA[BSB3] Value of offshore contracts

from US in 2003

$10 billion

II. Estimates of Workers Vulnerable to Offshoring

Bardhan and Kroll (2003) US workers in service jobs

vulnerable to offshoring


Progressive Policy Institute


US IT jobs vulnerable to



III. Projections of Offshoring

Wired magazine (Pink 2004) Service jobs leaving the United

States each year for the

foreseeable future


Prism (2004) meta-analysis Percentage of IT jobs lost from

US over next five years

7% to 8%

ITAA[BSB4] Value of offshore contracts

from US in 2008

$31 billion

Deloitte Research (2003

report cited in GAO 2004)

Financial services jobs that

may move offshore

850,000 (15% of

industry employment)

Forrester (2004) report cited

in GAO (2004)

U.S. service jobs lost by 2015 3,300,000

Goldman Sachs (2003 report

cited in GAO 2004)

Services and manufacturing

jobs lost over coming decade

Up to 6 million jobs

Evalueserve (2003) All jobs lost 2003-2010 1.3 million worst case

Evalueserve (2004A) Total jobs offshored in IT and

non-IT business process

operations (BPO) in 2010

775,000 IT jobs

1,414,000 non-IT BPO


Evalueserve (2004B) Growth of value of knowledge

process offshoring (KPO) from


From $1.29 billion in

2003 to $17.0 billion in

2010 (46% annual

growth rate)

Shaw quoted in McDougall


IT jobs moving offshore in 30


30% of IT jobs offshore

within 25-30 years

Gartner quoted in McDougall


Percent of U.S. IT jobs

offshored in 2005 and 2015

Will increase from 5% in

2005 to 30% in 2015

Two studies estimated the number of jobs in the United States that are vulnerable to

offshoring, and they found 12 to 14 million jobs could be offshored (Bardhan and Kroll

2003; Progressive Policy Institute 2004). Both studies note that their figures represent an

upper bound on offshoring activity that could occur not a projection of what will take place.

Thus, these studies are useful not so much for the numbers they provide as for identifying

sectors and occupations subject to offshoring. Bardhan and Kroll, for example, include in

their 14 million jobs vulnerable to offshoring office support, business and financial support,

Page 85

computer and mathematics professionals, paralegals and legal assistants, diagnostic

support service jobs, and medical transcriptionists.

A number of studies provide projections of offshoring activity in the United States. It is

difficult to make comparisons among the studies for several reasons. The major obstacle is

that the studies measure different sectors of the economy (e.g., IT, services,

manufacturing, business processing operations, knowledge process offshoring), use

different measures of the extent of offshoring (e.g., jobs lost, percent of jobs lost, value of

jobs or business lost), and different start and end points. Few of the studies provide details

on how they developed their projections which makes it difficult to assess the

reasonableness of the assumptions and the soundness of the methodology. The

Evalueserve studies provide more details on their methods so one can assess the

assumptions and methods better than for most other studies. For example, a critical

assumption in Evalueserve’s work is that there will be a large shortage of labor in the United

States and that a significant part of the solution will be from offshoring. Since Evalueserve

has described its assumptions and methods in some detail, other analysts can make their

own assessment of how reasonable these assumptions and methods are.

All the projections indicate that offshoring of service jobs in the United States in general

and of IT jobs specifically will continue to grow, but there is some disagreement about how

rapidly the growth will take place. Some of the studies project absolute numbers of workers

lost over a given number of years such as Forrester’s projection of 3.3 million service

workers by the year 2015. The problem with these projections is that most of them do not

give a baseline for understanding the significance of the job losses. They often do not tell

you the size of the population from which these projected losses will be taken. In fact, it is

a difficult task to count the number of IT or service workers in the United States. Here is

one part of a lengthy analysis of this issue from a Computing Research Association study

done in 2000 about the IT workforce.

Commerce used the narrow definition of the Bureau of Labor Statistics

classifications: computer scientists and engineers, systems analysts, and computer

programmers. The Information Technology Association of America (ITAA) used a

broader definition: any skilled worker who performs any function related to

information technology, which itself is defined as the "study, design, development,

implementation, support or management of computer-based information systems,

particularly software applications and computer hardware." (Freeman and Aspray


Not surprisingly, different definitions lead to different numbers. The Bureau of Labor

Statistics counted a little more than 2 million workers in 1997 and about 3 million today.

Between 2000 and 2004, the ITAA has counted between 10 and 11 million IT workers.

Service workers account for a very significant portion of the American labor force,

amounting to many tens of millions of workers. Some significant fraction of these service

workers are enabled by information technology, but it is hard to count how many and to

what extent.

Some of the other studies project in terms of the percentage of the jobs lost. These vary

significantly in the percentages they quote and the length of time over which this job loss

takes place. But perhaps more importantly, it is hard in some cases to know what a given

percentage projection means or to compare across these projections by different

organizations since their meanings are often different from one another. When a study says

that 25% of the US IT jobs will be lost by a given year, does it mean that the IT workforce

in the United States will be a quarter smaller than it is today? Does it mean that there will

be three IT jobs in the United States at that date in the future for every IT job in another

country doing work for a US firm? Are the numbers calculated on a base of the number of IT

Page 86

jobs today or at that time in the future? Would it matter if there were a sizable number of

IT jobs in other countries servicing the US economy if the number of IT jobs that remained

in the United States continued to grow?

Some of the projections are for specific slices of the offshoring market. For example,

Evalueserve (2004B) has projected a high rate of growth—46 percent annual compound

rate of growth—in what they term knowledge process offshoring (KPO). Evalueserve

distinguishes KPO from other business process operations (BPO) by the high level of

knowledge intensiveness required. Examples of opportunities in KPO include intellectual

property research; equity, financial, and insurance research; data search, integration, and

management; analytics; and research and information services in human resources

(Evalueserve 2004B, p. 4). What happens in the KPO slice that is developing from a small

base, making it easier to have high growth percentages, may well not be true for other

segments of the offshoring industry.

How do we put all this in perspective? Looking only at the number of jobs lost to

offshoring is a narrow and one-sided way of looking at the situation. If one wants to know

how many jobs there will be for IT workers in the United States in the future, one needs to

look not only at all the ways in which jobs are lost (including not only those moved offshore

but those lost through companies downsizing or going out of business) but also look at the

number of new jobs that are created in the IT occupations.

It is useful to consider what has happened over the past five years and compare that

situation with some of the projections just mentioned about IT job growth. The reports

discussed earlier in this section indicate that the United States has lost several hundreds of

thousands of jobs to offshoring since the height of the dot-com boom. Does that mean that

we have fewer jobs today in the United States in the IT field than we did then? The last

section of Chapter 1 contains an analysis of recent US Bureau of Labor Statistics numbers

(see Table 8). It shows that the number of US IT workers is actually higher today than in

1999, at the height of the dot-com boom, despite the hundreds of thousands of jobs lost to

offshoring. People who study the overall US labor market will not be surprised to learn that

the situation is not as dire as it is made out to be by those who dwell on the offshoring

losses. This is because history shows tremendous and continuing churn in the American

labor market with massive numbers of jobs lost and jobs created each year, but with a fairly

consistent pattern that the number of jobs created is larger than the number of jobs lost.

These patterns do not hold true for all occupations of course, we have significantly fewer

telephone operators than we once had, for example, but the Bureau of Labor Statistics in its

ten-year forecasts continue to believe that the IT occupations will experience overall

sustained growth and, in fact, several IT occupations will be among the fastest growing

occupations in the next decade. The ITAA study mentioned previously also suggests that

the number of IT jobs created in the United States in the future will be robust.

One can similarly ask about the effect of offshoring not on jobs but on wages for US IT

workers. The same Bureau of Labor Statistics data just cited indicates that, since the height

of the dot-com boom and throughout the dot-com crash, even through a recession in the US

economy, IT wages have continued to rise at about twice the percentage of inflation.

Wages did not rise at the same rate in all IT occupations. High-skill jobs rose at the highest

rates, for example, computer science research salaries rose at 5.6% per year and computer

systems manager salaries at 5.7% per year. Wages for low-skills jobs rose less rapidly.

For instance, computer support personnel wages rose at only 2.1% per year, fairly close to

the rate of inflation during this time. Even the much talked about programmer, whose job

is expected to be particularly vulnerable to offshoring, had wages rise by a healthy 3.7%

per year.

Page 87

What will happen in the future is hard to predict. If the United States remains innovative

in the IT field and if this innovation continues to have an important positive influence on the

US economy as was the case throughout the 1990s, then it is likely that the United States

will continue to create large numbers of jobs for IT workers. If the country does not

continue to promote innovation and it cedes large portions of its IT industry to other

countries, then the future is likely to be much bleaker for American IT workers.

What Does the Data Tell Us About the Size and Impact of Offshoring in Countries Other

Than the United States That Offshore Work?


The European Union is the largest offshorer of IT software and services after the United

States. In one project of the European Commission’s Information Society Technologies

Programme, known as the STILE project, there was an effort to cull data on the ITproducing

sector in Europe. The study group tried to use the accepted classification

scheme, Nomenclature Generale des Activites Economiques dans les Communautes

Europeenes (NACE) (General Industrial Classification of Economic Activities within the

European Communities). They found that this was not particularly useful because IT

activities are in many cases bundled with non-IT activities in the classification codes, and it

was impossible to separate out services from products. More generally, there seems to be

very little data, if any, collected by government organizations that provides information

about the state of offshoring in Europe. As one major survey of the literature on European

offshoring summarizes:

It is important to emphasize, along with the OECD (van Welsum, 2004) that there

are currently no reliable statistical indicators of the extent or nature of global

outsourcing. It is not possible, either through the trade statistics or the EU

occupational and employment statistics, to track statistics of imports and exports of

business services to identify with any accuracy which components of these services

represent jobs.

For evidence we must therefore look to the results of market research, one-off

surveys and case studies or anecdotal evidence. It must be emphasized that these

vary in their reliability and their conclusions. They may be coloured, either positively

or negatively, by the specific interests of the agencies who commissioned them,

depending on their point of view. In between, lie many analyses which strain for

objectivity but are hampered by the lack of solid empirical evidence….

An even greater confusion exists in the EU [than in the United States] where there

are notably few academic, systematically led investigations in the area of European

and offshore outsourcing. Estimates of the impact on Europe are vague, especially in

relation to outsourcing to smaller Asian countries and eastern European states.

(Huws et. al. 2004, p. 10)

Table 2 provides an overview of some of the consultant data on the extent and impact of

offshoring in Europe. The amount of offshoring is clearly at substantially lower levels than

in the United States. These reports indicate that only about 5% of European businesses are

currently offshoring, and not quite a third of European companies are even planning for

offshoring. Huws et. al. (2004) point out that most studies agree that 2 to 3% of all EU

service employment will be lost to offshoring by 2015 but notes that this loss is likely to be

less than the number of new jobs created so that the European service industry is likely to

expand its employment over time.

It is true today that Europe overall is not losing IT service jobs. Even at the country level,

this is mostly true, and where there are national losses, in every country other than

Denmark, any losses in computer jobs have been more than offset by growth in business

Page 88

service jobs (Eurostat, Community Labour Force Survey). Other than the United Kingdom,

Western Europe has a decided preference for selecting Eastern European firms to do their

offshore work (Pierre Audoin Consultants (2003), quoted in Huws et. al. 2004).

Nevertheless, there is a sizable amount of work sent from Germany as well as from the

United Kingdom to India. Eurostat data (Community Labour Force Survey) shows that the

EU countries that are growing most rapidly in IT business activities are the new member

states such as Hungary which had been behind the rest of the EU in percentage of IT

activities. These data are consistent with, but not proof of, these countries becoming prime

destinations for European offshoring.

Table 2-2: The Extent and Impact of Offshoring on Europe

Forrester (August 2004 as

quoted in Gumbel 2004 and

Knapp 2004)

Jobs lost by Europe to offshoring by


1.2 million,


150,000 IT


Deloitte Research (quoted in

Matlack et al. 2004)

Financial-services and high-tech jobs

that will migrate from Europe to lowwage

countries by 2008


Heidrick and

Struggles/PeopleSoft (2004) as

reported in van Breek (2004)

CEOs of European companies who

reported offshoring work to low-cost

labor markets as a key business issue


European Commission 2000 (as

quoted in Huws et al. 2004)

European companies offshoring 5.3%

Forrester Research (as quoted in

Huws et al. 2004)

Value of offshore spending in western

Europe in 2004 and 2009

1.1 billion, 3.6

billion Euros

Gartner (as quoted in Huws et

al. 2004)

European business expected to

include offshoring in their business

plans by end of 2005


Huws et al. (2004) EU service employment lost to

offshoring by 2015

2% to 3%


The United Kingdom is the largest offshorer in Europe by a considerable margin, and this

trend is likely to continue at least for the next few years. Forrester Research estimates that

the United Kingdom will be responsible for three-quarters of all European offshoring in five

years (Huws et. al. 2004). 61% of UK companies send IT work across national boundaries

compared to only 15% in Germany, the second largest offshorer in Europe (Roland Berger

2004). Amicus (2004) claims that a thousand UK jobs are being lost to offshoring each

week and projects a loss of 200,000 UK IT jobs by 2008 with a significant number of these

positions in IT support.[BSB5] Management Consultancies Association predicts 25% growth

in offshoring of call centers between 2003 and 2008 (Huws et. al. 2004). The

Communications Workers Union (2004) predicts almost 200,000 call center jobs moved

from the United Kingdom to low-cost countries between 2004 and 2006.

According to the consulting firm Evalueserve (2004), the United Kingdom will face a

shortage of 714,000 workers by 2010 due to the aging of the workforce. IT is one of the

fields projected to have the greatest shortages. Evalueserve estimates that 342,000 of

these 714,000 jobs can be filled by immigrants and notes that the remainder will need to be

Page 89

filled in other ways if the country is to avoid a decline in GDP due to an insufficient labor

pool. Evalueserve claims that, by the end of 2002, 31,100 jobs in the service sector in the

United Kingdom had been moved offshore, and the company projects that 272,000 more

jobs will move offshore in the period 2003-2010. Focusing more narrowly on IT and

software development jobs only, Evalueserve claims that 18,000 jobs had been moved

offshore by the end of 2002 and 84,000 more would be moved between 2003 and 2010.


In Germany, there are no federal statistics that help to track the number of jobs offshored

(Bartsch 2004). For various reasons, it is expected that offshoring will not be as significant

in Germany as in the United States: few people in low-wage countries speak German,

Germany has fewer low-skill jobs than the United States of the sort that have been

offshored, stricter layoff regulations make it more costly in Germany than in the United

States to lay off workers and relocate the work overseas, and there has been less political

fallout over jobless recovery after the last recession than in the United States. Germany

historically does not have such strong labor rebounds after a recession as the United States.

After the most recent recession, job recovery was close to the historical norms in Germany,

while in the United States the recovery led to an unprecedented small number of jobs


Using an analysis of German foreign direct investment in Central and Eastern Europe by

the Oesteuropa-Institut, together with employment trends in German foreign affiliates and

correction factors for the German foreign affiliates numbers (because they underreport

activities by small and medium-sized companies), Elga Bartsch arrived at an estimate on

behalf of the consulting firm Morgan Stanley that the number of jobs (of all kinds) offshored

to Central and Eastern Europe over the past decade is approximately 600,000. She also

considered an alternate source, the European Restructuring Monitor (ERM), which “compiles

information on major corporate restructuring announcements from daily press runs of the

major national newspapers in the European Union and classifies them by country, industry,

and reason for the restructuring” (Bartsch 2004). The ERM found that 117,000 jobs losses

were announced in Germany from January 2002 into mid-2004, and that 3% were lost to

relocation of production facilities and another 0.3% to outsourcing. In another study

commissioned by the Deutsche Bank, it was estimated that 50,000 German IT-related jobs

had been relocated outside Germany up until the time of the publication of the study (Huws

and Flecker 2004).

In a McKinsey study (Farrell 2004), Germany shows only a $0.80 return on each dollar

invested in offshoring, compared to $1.14 return on the dollar invested in offshoring for the

United States. There are multiple reasons for this: German companies have higher

management costs because of language and cultural issues; offshoring work is frequently

contracted to Eastern Europe where the wages and infrastructure costs are higher than in

India; Germany is not able to capture much in high-tech exports through offshoring because

of the dominance of US firms in these export markets; Germany gains practically nothing

from repatriated earnings (i.e., from offshoring providers abroad that German companies

have invested in); and most important of all, as many as 60% of German workers have

difficulty being re-employed when they lose their jobs through offshoring which is a major

drag on the German economy.

What Does the Data Tell Us About the Size and Impact of Offshoring in India?

By far, the most complete data about the Indian software industry is provided by the

Indian trade association, NASSCOM. Table 3 provides a snapshot of the offshore industry

based on the NASSCOM Strategic Review 2005. You can see that the industry is vibrant

with growth in the 20 to 30 percent range each year. IT software and services are still the

largest export, far ahead of hardware and ITES/BPO, but the IT-enabled services are

Page 90

growing more rapidly than any other sector of the Indian offshoring industry. The United

States, followed distantly by the United Kingdom, is the destination of most of this work.3

About 60% of the ITES/BPO work is conducted by multinationals with this percentage

edging higher over time. The export trade is growing much more rapidly than the domestic

IT/ITES market in India, from 58% to 78% by revenue from 1996 to 2003 (RIS 2004 as

quoted in UNCTAD 2004). The software and service industry is becoming increasingly

important to India’s economy, growing from 3% to 21% of India’s total exports between

1996 and 2003 (RIS 2004 as quoted in UNCTAD 2004). Although there are more than 3000

software and service exporters in India, 25 of these firms collect more than 60% of the

revenue (Prism 2004).

Table 2-3: The Extent and Impact of Offshoring in India

Indian national economic output attributable to the ITITES

industry in 1997-98


Indian national economic output attributable to the ITITES

industry in 2003-04


Indian IT industry revenue in IT services and software in



Indian IT industry in hardware in 2003-04 22%

Indian IT industry in ITES/BPO in 2003-04 18%

Growth in Indian IT and ITES exports $9.9 billion in 2002-03

$$13.3 billion in 2003-04

IT and ITES export from India to the United States in



IT and ITES export from India to Europe in 2003-04 22% (mostly to the United


India IT services and software revenue in 2003-04 $12.8 billion (29.6% increase

over previous year)

Jobs added to Indian IT services and software sector in


98,000 (more to domestic

companies than multinationals)

Growth in Indian IT services revenue $1.9 billion in 2002-03

$2.5 billion in 2003-04

Growth in Indian offshore product development exports

(includes exports of software products made by Indian


$560 million in 2002-03

$710 million in 2003-04

(mostly produced by

multinational captives)

Multinational companies that opened captive centers in

India since early 2001


3 Joseph and Parayil (2004 as quoted in UNCTAD 2004) claims 80%. Table 3 gives 69% for IT and

ITES combined.

Page 91

Value of research conducted by Indian captives of

multinationals in 2003-04

$800 million to $1 billion

Value of research conducted by Indian captives of

multinationals in 2008 (projected)

$11 billion

Call center seats 96,000 in 2003; 158,000 in


Source: NASSCOM Strategic Review 2005

What Does the Data Tell Us About the Size and Impact of Offshoring in Countries Other

Than India That Do Software Work for Export?

Quantitative information about software exports especially those related to offshoring is

difficult to locate for most countries. Table 4 provides a smattering of data that we have

been able to locate although we cannot attest to its correctness.

Table 2-4: Nations Other than India with Offshoring Industries

Country Statistic Source

Australia $21B commercial service exports in


(22% computer and communications)

World Development Index database

call center seats: 135,000 in 2003,

146,000 in 2004

Barbados $1.1B commercial service exports in


(16% computer and communications)

World Development Index database

Belarus $1.5B commercial service

exports in 2003

(24% computer and communications)

World Development Index database

Brazil $9.6B commercial service exports in


(50% computer and communications)

World Development Index database

Canada HRDC estimates 500,000 Canadians

work in call centers

Prism (2004)

Cape Verde


$211M commercial service exports in


(9% computer and communications)

World Development Index database

China 8,000 software and service providers,

_ of whom have fewer than 50

employees and only five have more

than 2,000 employees. (Yuan 2005)

McKinsey (January 2005)

Page 92

$1.5B in 2003 Prism (2004)

Call center seats: 38,000 in 2003;

54,000 in 2004



$7.8B commercial service exports in


(24% computer and communications)

World Development Index database



$3.4B commercial service exports in


(5% computer and communications)

World Development Index database

Ghana $612M commercial service exports in


(11% computer and communications)

World Development Index database

Guatemala $954M commercial service exports in


(19% computer and communications)

World Development Index database

Hong Kong Call center seats: 10,000 in 2003;

10,700 in 2004

Hungary $7.9B commercial service exports in


(41% computer and communications)

World Development Index database

Ireland leads the global market in offshore IT

services with 25% of market

World Investment Report 2004

$3.8B in 2000, 8.5B Euros in 2003 Prism

60% of EU software is developed or

localized in Ireland; software industry

is 11% of GDP

55% of Ireland’s 28,000 IT

professionals are employed by

multinationals and account for 90%

of Irish software exports

$38B commercial service exports in


(61% computer and communications)

World Development Index database

Israel $1.9B value in offshoring exports in


Prism (2004)

Page 93

Latvia $1.5B commercial service exports in


(19% in computer and


World Development Index database

Madagascar $202M commercial service exports in


(32% computer and communications)

World Development Index database

Malaysia call centers growing at between 100

and 200% per year since 2000

World Investment Report 2004

$14B commercial service exports in


(33% computer and communications)

World Development Index database

Mauritius $1.3B commercial service exports in


(17% computer and communications)

World Development Index database

Mexico $13B commercial service exports in


(7% computer and communications)

World Development Index database

Morocco $5.1B commercial service exports in


(18% computer and communications)

World Development Index database

Call center seats in all North Africa in

2005 (Morocco has largest share):



New Zealand Call center seats: 12,000 in 2003;

13,500 in 2004

Philippines 27,000 people in call center jobs in

2003 and growing rapidly

World Investment Report 2004

$250 software, or $1B including BPO


Prism (2004)

Call center seats: 20,000 in 2003;

40,000 in 2004

Poland number of jobs in BPO will increase

from 3,000 in 2004 to 200,000 in


McKinsey & Co. (as quoted in

Wagstyl 2004)

Romania $3B commercial service exports in


(42% computer and communications)

World Development Index database

Page 94

Russia $150 – 200M value of offshoring

exportsin 2003

Prism (2004)

Senegal $3890M commercial service exports

in 2002

(40% computer and communications)

World Development Index database

Singapore Call center seats: 10,000 in 2003;

10,100 in 2004



$3.3B commercial service exports in


(28% computer and communications)

World Development Index database

South Africa number of call centers to expand

from 494 in 2004 to 939 in 2008);

serving the English but also the

German populations.

Datamonitor (as quoted in

Chatterjee 2004)

Employees working in call centers in


New York Times (Feb 2, 2005)

$6.4B commercial service export in


(9% computer and communications)

World Development Index database

Thailand Call center seats: 11,000 in 2003;

13,000 in 2004

Tunisia $2.8B commercial service exports in


(16% computer and communications)

World Development Index database

Ukraine $5B commercial service exports in


(11% computer and communications)

World Development Index database

2.3 Conclusions

Even in the face of offshoring, economists generally continue to believe in the theory of

comparative advantage, that if each country specializes in the production of goods where it

has comparative advantage and trade is not restricted, both countries can enjoy greater

total consumption and well being by trading with one another. Some economists, notably

Gomory and Baumol, have pointed out that it is possible for a country to lose under free

trade. In the short-term, the question is one of jobs and wages. Are the jobs lost to

offshoring in developed countries compensated for by new job creation in these countries

which might come, for example, from the lower cost of development and production, faster

development time, higher quality, or round-the-clock service associated with using an

offshore workforce to supplement or supplant the domestic workforce? Similar questions

can be asked about wage rates. The analysis by Mann of hardware offshoring to Asia in the

1990s suggests by analogy that it is entirely possible for a developed nation to be much

better off through offshoring of its software work. Recent Bureau of Labor Statistics shows

that IT jobs and wages have generally increased in the United States from the height of the

Page 95

dot-com boom until late 2003, a period during which there was active offshoring. However,

lack of data limits what we can say about the impact of offshoring.

Leaving aside for the moment the question of whether offshoring has hurt aggregate jobs

or wages, there are at least two ways in which offshoring might harm a developed country.

Offshoring clearly can do harm to individuals who lose their jobs through offhsoring and to

local communities that lose large numbers of jobs or particular businesses through this type

of trade. The general sense among economists is that the only solution to this is a political

one of providing a safety net to workers. This might include wage insurance, extended

unemployment benefits, retraining, and perhaps others benefits. However, the costs of

providing this safety net are great. The other way that offshoring can harm a developed

country in the long run is to erode the country’s capability to innovate. As a developed

country loses its edge in innovation, it becomes less able to remain competitive. Enabling a

country to remain innovative is a matter of education, research funding, and immigration

policy. These issues are discussed in Chapters 7 and 8.

What does the available data tell us about the extent and trajectories of offshoring? First,

considering the United States which has seen the largest amount of offshoring and is

subject to the greatest amount of loss through offshoring, we know that there are perhaps

12 to 14 million jobs vulnerable to offshoring. However, this number represents a high

upper bound on potential job losses, and nobody believes that all these jobs will be lost. So

far, annual job losses have probably been no more than 2 to 3% of the IT workforce in the

United States. Moreover, there are questions about the reliability of these numbers because

of the definitions and other methodological issues. The meaningfulness is also at question

because these job losses have to be placed in perspective with the much higher level of job

loss and creation that occurs every year in the United States and, in particular, with the jobs

that are created directly or indirectly because of companies sending work offshore. The

data simply does not exist that would enable a full analysis of the impact of offshoring on

the US IT workforce. Anecdotal information, together with data, suggest that the United

States will continue to increase the amount of work it offshores at double-digit percentage

rates at least for the next few years. Programming and related technical work continues to

be the work most likely to be offshored, but IT-enabled services are rapidly taking a greater

share. There is also rapid growth, from a small base, in the offshoring of higher-value

activities such as knowledge processing and research.

Data about countries other than the United States or worldwide data are much harder to

come by than data for the United States. There are reports of offshoring industries or

sectors of these industries that have grown in India, China, and a few other countries by 20,

30, or higher percentages per year with projections that these growth rates will continue for

varying lengths of time into the future. It is difficult to evaluate these projections but it

seems likely that there will be continued rapid growth at least for the next few years.

Whether these growth rates will be 10, 20, 30, or 40 percent per year is beyond our ability

to project based on existing data. It appears that India will continue to be the primary

destination of offshoring, with China growing rapidly. However, Chinese software activities

are devoted to a significant extent on the emerging domestic market not the export market.

There appears to be some promise of growth for the main nearsourcing countries such as

Canada and those in Eastern Europe despite the fact that their wage rates are higher than

those in the low-wage Asian countries. There are at least limited opportunities for

offshoring work by companies located in Africa, Latin America, and low-wage Asian

countries other than India and China. Data, together with anecdotal evidence, suggests that

Western Europe is beginning to increase the amount of work it offshores. The United

Kingdom has been by far the largest offshorer in Europe, and this is likely to continue to be

true for the next few years. And Germany, in particular, has begun to increase the amount

Page 96

of work being offshored in the past year or two. Japan is set to increase the amount of

work offshored especially to China.

There are numerous problems with the current state of data. Definitions used in reporting

offshoring’s growth and impact are inconsistent with one another. All of the obvious metrics

that could be used to measure offshoring have limitations. Government statistical

organizations, such as the US Bureau of Labor Statistics and the Bureau of Economic

Analysis, provide the greatest promise in providing good data because of their highly trained

staffs and long traditions of quality, reliability, and objectivity. Governments collect data,

however, in connection with existing policy issues, and the offshoring phenomena is

sufficiently different that existing government data sources turn out to be not very useful.

Trade organizations and consulting firms are not disinterested parties, and these

organizations are often unwilling to make public the methods and assumptions by which

they arrive at their results so it is not surprising that there is some skepticism in the

economic community about the credibility of their results. For many parts of the world,

little or no data is being gathered.

A professional society such as ACM itself is not in a good position to collect data.

However, it can encourage the principal data gatherers – governments, trade associations,

and consultants – to improve their offshoring data practices. We need clear definitions,

careful choice of metrics, data that separates the impact of offshoring on job loss from other

causes of job loss such as business cycles and technological change, and data that

measures the various aspects of offshoring (jobs created, gains in wealth to companies and

nations, impact on wage rates, etc.) not just job loss. Thus, we recommend that the

following steps be taken:

Standard definitions of offshoring and related terms should be developed. A good starting

place is the diagram developed by the US General Accountability Office (Figure 2) that

provides a complete description of all outsourcing and offshoring activities. These definitions

should be used by all countries participating in the global software market, not just the

United States.

The US Department of Labor should gather data on layoffs that is more suited to

measuring offshoring than the current Mass Layoffs Statistics data. Ideally, such data

should be collected on a regular basis, but even a one-time special effort would be useful.

Other countries should collect similar data.

The United States should improve the collection of data on imports and exports of

services by country, following the guidelines recommended by the General Accounting Office

(2004). Other countries should follow similar practices, and practices should be consistent

from country to country.

Data on direct investment abroad by source country and multinational company

operations should be improved also following the recommendations of the GAO (2004).

R&D and design activities in low-wage countries should be tracked especially in affiliates

of multinational firms.

All organizations, private as well as public, that are creating statistical information about

offshoring should be transparent about their methods and assumptions.

Developing better current data and adopting standardized definitions should help to

improve projections of offshoring. We are not sanguine, however, about the likelihood of

developing good projections any time soon. In the United States, the BLS has been

modestly successful in developing ten-year occupational projections, but projecting the size

and effects of offshoring appears to be more difficult.

Page 97

Figure 2-2

2.4 Bibliography

Alpert, A. and Auyer, J. 2003. Evaluating the 1988-2000 Employment Projections. Monthly

Labor Review (Oct.) 13-37.

Amicus MSF. 2004. Back to the Future as Call Centres Go Same Way as Manufacturing.

London, UK.

Associated Press 2005. 400K+ High-Tech Jobs Lost. (Sept. 15) Seattle, WA. (As reproduced

on April 4, 2005.)

Atkinson, R. 2004. Understanding the Offshoring Challenge. Progressive Policy Institute


Bajpai, N., Sacks, J.D., Arora, R., and Khurana, H. 2004. Global Services Sourcing: Issues

of Cost and Quality. Center on Globalization and Sustainable Development. Working Paper

No. 16. Columbia University, New York, NY.

Bardhan, A.D. and Kroll, C.A. 2003. The New Wave of Outsourcing. Fisher Center for Real

Estate and Urban Economics, University of California, Berkeley, CA. (Fall).

Bartsch, E. 2004. Germany: Offshoring – More a Myth Than a Matter? Morgan Stanley

Global Economic Forum (Aug.).

wed.html and

Bednarzik, R.W. 2005. Restructuring Information Technology: Is Offshoring a Concern?

Monthly Labor Review (Aug.) 11-21.

Page 98

Bhagwati, P. and Sriniviasan, T.N. 2004. The Muddle Over Outsourcing. Journal of Economic

Perspectives, 18, 4, 93-114.

Bishop, J.H. and Carter, S.D. 1991. How Accurate Are Recent BLS Occupational Projections?

Monthly Labor Review (Oct.) 37-43.

BIVENS, J., 2004. Will the New International Division of White-Collar Work Make the U.S.

Rich? ACM Job Migration Task Force Meeting (Dec.) Washington, DC.

Bronfenbrenner, K. and Luce, S. 2004. The Changing Nature of Corporate Global Restructuring: The

Impact of Production Shifts on Jobs in the US, China, and Around the Globe. Corporate Restructuring

and Global Capital Mobility (Oct.).

Business Week Online. 2005. Just the Bright Side, Thanks. (Oct. 17).

Chatterjee, S. 2004. South Africa Seen as India’s Rival in Outsourcing. Indo-Asian News

Service (Nov. 21). (As appeared in Yahoo! India News, Nov. 28, 2004.)

Communications Workers Union. 2004. The Threat of Outsourcing UK Call Centre Jobs

Offshore – the Current Situation and Future Developments. London, UK.

Cusumano, M. 1991. Japan’s Software Factories. Oxford University Press.

Dossani, R. 2005. IT Services Offshoring to India: India’s Position in the Supply Chain. ACM

Job Migration Task Force Meeting (March) Palo Alto, CA.

Egger, H. and Egger, P. 2001. International Outsourcing and the Productivity of Low-skilled

Labor in the EU. WIFO (Australian Institute of Economic Research). Working Paper 152.



Egger, H. and Egger, P. 2003. Outsourcing and Skill-Specific Employment in a Small

Economy: Austria after the Fall of the Iron Curtain. Oxford Economic Papers 55, 4, 625-643.

Egger, P., Pfaffermayr, M., and Wolfmayr-Schnitzer, Y. 2001. The International

Fragmentation of Austrian Manufacturing: The Effects of Outsourcing on Productivity and

Wages. North American Journal of Economics and Finance 12, 257-272.

Economic Policy Institute. 2004. Offshoring Frequently Asked Questions. (June).

Evalueserve. 2003. The Impact of Global Sourcing on the US Economy. (Oct. 9).

Evalueserve. 2003. Impact of Global Sourcing on the UK Economy 2003-2010.

Evalueserve. 2004. The Next Big Opportunity—Moving up the Value Chain from BPO to KPO.

(July 13).

Evalueserve. 2004. Offshoring of IT Services—Present and Future. (July 13).

Farrell, D. 2004. Can Germany Win from Offshoring? McKinsey Global Institute (July).

Feenstra, R.C. and Hanson, G.H. 2001. Global Production Sharing and Rising Inequality: A

Survey of Trade and Wages. National Bureau of Economic Research. Working Paper 8372.

Fielding, R. 2003. Enlarged EU to Attract Offshoring. (Oct. 3).

Freeman, P. and Aspray, W. 1999. The Supply of Informatic Technology Workers in the

United States. Computing Research Association. Washington, DC.

Goldman Sachs. 2004. Offshoring by the Numbers (May). (As quoted in the Center for

American Progress.)

Gomory, R. E. and Baumol, W.J. 2000. Global Trade and Conflicting National Interests. MIT

Press, Cambridge MA.

Page 99

Gorg, H. and Hanley, A. 2004. International Outsourcing and Productivity: Evidence from

Plant Level Data. University of Nottingham.

Gumbel, P. 2004. Au Revoir, Les Jobs,. Time Europe (Online edition, Oct. 3).

Huws, U. and Flecker, J. 2004. Asian Emergence: The World’s Back Office? Institute for

Employment Studies. Report 409. Brighton, UK.

Huws, U., Dahlmann, S., and Flecker, J. 2004. Outsourcing of ICT and Related Services in

the EU. European Foundation for the Improvement of Living and Working Conditions.

Dublin, Ireland. Available at

Kletzer, L. 2004. Trade-Related Job Loss and Wage Insurance: A Synthetic Review. Review

of International Economics 12, 5 (Nov.).

Knapp, P. 2004. 1.2 Million European Jobs to Flee Offshore. (Aug. 24).

Lamoreaux, N. and Sokoloff, K. 1996. Long-Term Change in the Organization of Inventive

Activity. Proceedings of the National Academy of Science 93, 12686-12692.

Lamoreaux, N. and Sokoloff, K. 1997. Location and Technological Change in the American

Glass Industry During the Late Nineteenth and Early Twentieth Centuries. National Bureau

of Economic Research. Working Paper 5938. Cambridge MA.

Mariani, M. 2001. Next to Production or to Technological Clusters? The Economics and

Management of R&D Location. Journal of Management and Governance 6, 2, 131-152.

Mann, C.L. 2003. Globalization of IT Services and White Collar Jobs: The Next Wave of

Productivity Growth. Institute of International Economics. International Economics Policy

Briefs 3-11 (Dec.).

Mann, C.L. 2004. What Global Sourcing Means for U.S. IT Workers and for the U.S.

Economy. Virtual Machines 2, 5 (July/August).

Matlack, C., Kripalani, M., Fairlamb, D., Reed, S., Edmondson, G., and Reinhardt, A. Job

Exports: Europe’s Turn. BusinessWeek Online (April 19).

McCarthy, J.C. 2004. Near-Term Growth of Offshoring Accelerating. Forrester Research


McDougall, P. 2004. Offshore Outsourcing Revenue Soaring at 20%-A-Year Pace. (Oct. 16).


McDougal, P. 2005. Exclusive: Gartner Predicts Increase in Offshore Outsourcing by 2015.

CRN (March).


McKinsey & Co. 2003. India Information Technology/Business Process Offshoring Case


McKinsey & Co. 2005. Can Germany Win From Offshoring?


Meta Group. 2004. Annual IT Staffing and Compensation Guide.

NASSCOM. 2004. Strategic Review 2004: The IT Industry in India (Feb.) New Delhi, India.

Pink, D.H. 2004. The New Face of the Silicon Age. Wired (Feb. 12).

Prism Economics and Analysis. 2004. Trends in the Offshoring of IT Jobs. Software Human

Resources Council. Ottawa, Canada.

RIS. 2004. European Foundation for the Improvement of Living and

Working Conditions. (Full citation in bibliography of World Investment Report 2004.)

Page 100

Roland Berger Strategy Consultants. 2004. Global Footprint Design: Mastering the Rules of

International Value Creation. Munich, Germany.

Samuelson, P. 2004. Where Ricardo and Mill Rebut and Confirm Arguments of Mainstream

Economists Supporting Globalization. Journal of Economic Perspectives 18, 3, 135-136.

Sasaki, T. 2004. Upgrading of IT Based Services: Trends, Challenges, and Policy

Implications. The UNCTAD-ASEAN Seminar on Service FDI and Competitiveness in Asia.

Kyoto, Japan. (March).

Scholl, R.S., Sinha, D., Datar, R., and Chohan, S. 2003. India Will Generate $13.8 Billion

from Offshore BPO Exports in 2007. Gartner Dataquest Report (June) Stamford, CT.

Seeley, R. 2003. Report: 150,000 U.S. Software Jobs Lost Last Year. (Dec. 2).

Sutthiphisal, D. 2004. The Geography of Invention in High- and Low-Technology Industries:

Evidence from the Second Industrial Revolution. UCLA, Department of Economics. Working

paper. 2004. World on Brink of Surge in Offshore Provision of Services: U.N.

(Sept. 22).

United Nations Conference on Trade and Development. 2004. The Offshoring of Corporate

Service Functions: The Next Global Shift? World Investment Report 2004 (Chapter 4).

United Nations, New York, NY and Geneva, Switzerland.

U.S. Government Accountabillity Office. 2004. International Trade: Current Government

Data Provide Limited Insight into Offshoring of Services. GAO-04-932, (Sept.).

Van Breek, Y. 2005. Offshoring a Top Priority for European CEOs. The Conference Board.

2005. (Online publicity for The Conference Board, Report 1353-04-RR. CEO Challenge


Wagstyl, S. 2004. Budapest, the Next Bangalore? New EU Members Join the Outsourcing

Race. The Financial Times (Sept. 21). (As reprinted in YaleGlobal Online.)

Yuan, L. 2005. Chinese Companies Vie for a Role in U.S. IT Outsourcing. Wall Street Journal

Online (April 5).

*M. Blasgen and M. Kenney took a leadership position in the writing of Chapter 3. Page 101

Authors: Govindasamy Balatchandirane, Michael Blasgen*, Arndt Bode, Charles H. House,

Martin Kenney*, Vivek Mansingh, Goran Marklund, Bankim Shah, Takashi Umezawa

Chapter 3: The Country Perspective

3.1 Introduction

This chapter examines the development and current status of the largest and most

dynamic nations in the evolving global software market with particular attention given to

offshoring. The focus is on software services and to a lesser extent on software products. In

particular, the chapter examines three pairs of regions that have special offshoring

relationships with one another: the United States and India, Japan and China, and Western

Europe and Eastern Europe/Russia. The selection of these three pairs is not meant to argue

that there are not other linkages, for example, between Western Europe and India; there

are. Western Europe is not monolithic in its offshoring patterns; it has different

geographical patterns, largely based on language capabilities. For example, the United

Kingdom sends its offshore work primarily to India, whereas Germany has strong

relationships with both Eastern Europe/Russia and India (see Figure 1). Thus the portrayal

here of national/regional pairs of offshoring partners, while representing some important

aspects of the global software industry, is clearly a simplification of an extremely

complicated map of offshoring.

Figure 3-1. Outsourcing from Germany

Source: Compilation from various newspaper articles by Martin Wildemann 2005:19

The decision to focus on only these regions and nations means that we omit some

countries that are active in the global trade in software. In particular, we touch only briefly

on Ireland and Israel who were among the earliest countries to enter the global market.











Other 1%



Eastern Europe

(in EU)

Eastern Europe (not

in EU) Asia (except China

and India)

Page 102

Although their software export markets are significant, they are relatively small and are not

expected to grow much especially in comparison, for instance, to India or China. Ireland

and Israel also do not appear to have a major effect on the global division of software labor.

We also omit many nations that have smaller software services export businesses such as

the Philippines and Mexico. Mexico, which in 2003 is estimated to have exported $30

million of software and software services (Singh 2003), is discussed in Chapter 4 through an

examination of one of its leading software export firms, Softtek. Mexico’s limited amount of

software services exports is typical of many developing nations. Rather than listing all of the

nations and the size of their industries (see some data on this topic in Chapter 2), this

chapter focuses on the most important ones, but suggests that they are not unusual, just

indicative of the larger pattern of globalization.

Offshoring has a long history and continues to evolve. National IT industries and

international supplier-customer relationships are part of a co-evolutionary process involving

many parties. Factors include government support, education, infrastructure,

telecommunications policy, finance, and even national perceptions. These factors interact

and gradually produce an environment more or less favorable to accepting relocated work.

This is exemplified in the case of China and its manufacturing sector where a sophisticated

manufacturing support infrastructure evolved over the past two decades to facilitate the

manufacturing of goods for export. In the case of India, higher educational institutions,

infrastructure, labor force, and government policies have evolved in a way that encourage

and support the production of IT services for the global economy. Each of the nations and

international relationships examined in this section is a product of similar co-evolution.

For firms, the decision of whether and where to offshore a certain business function

involves a complex calculation that balances a variety of concerns that include labor force

availability, government policy, factor costs, various kinds of risk, and comfort level with the

location. For each of these nations, this chapter examines both their past experiences with

offshoring and their current situation. We also consider their prospects for future growth.

3.2 The History of Software Offshoring

The origins of software offshoring are difficult to determine because large multinationals

such as IBM have long had overseas R&D facilities that were conducting software

development for the company’s global operations at the same time that they were

undertaking localization work for their domestic markets. The nations that first emerged as

software development sites for the global economy, that is, not for the domestic market,

were Israel and Ireland. Notice the distinction made here between undertaking software

development for the domestic market, which includes localization and even some

development based on unique features of the local market such as different accounting or

legal systems, as opposed to producing for the external market. Production for the local

market can displace jobs for workers in high-wage countries. However, it is unremarkable

that localization would be undertaken in the local market where knowledge of the language

and the specifics of the business culture and legal environment are the greatest. With

respect to job loss, there is far greater concern about the displacement of labor by a nation

producing for another nation’s market than for its own.

The United States is the overwhelming leader in the world software industry as the home

to such firms as Accenture, IBM, Microsoft, and Oracle. The only firms that rival these giants

are SAP (headquartered in Germany) in packaged software and Siemens Business Systems

(Germany) and Cap Gemini (France) in software services. Siemens Business Systems is

losing money and may be sold.

Page 103

Let us turn briefly to the small economies, Ireland and Israel, which pioneered software

and software services production for the global economy. The Israeli IT industry first

emerged in the late 1960s through an excellent educational system, military research, a

strong relationship to the United States based on geopolitics, and investment by

multinationals (de Fontenay and Carmel 2004). In the 1980s, Israelis began to found new

technology firms, many of which specialized in packaged software especially for security.

These Israeli firms often had cutting-edge technologies and, as part of their life cycle, very

soon established operations in the United States. The most successful of them listed their

stock on NASDAQ. Many of these firms were successful, but because of the small size of the

Israeli software industry their success is unlikely to lead to a massive relocation of

employment from the developed nations to Israel. The total Israeli employment in IT

services, including software, was 92,000 in 2000 or approximately 4 percent of the total

Israeli workforce (de Fontenay and Carmel 2004, 43). The Israeli Export and International

Cooperation Institute (2005) reports Israel as having 13,000 software professionals in 2002,

down from an all-time high of 14,500 in 2000. Whichever number is correct, Israel has a

much smaller software workforce than larger nations, though it is unusually large in terms

of the percentage of its own population. Even if Israel were to double its IT services

employment to an unheard of 8 percent of its workforce – amounting to some 200,000

employees – it would still be less than one-third the size of India’s 697,000 employees in

software and software-related services. Israel’s niche in the global software industry is as a

center of entrepreneurship at the highest technological level. It draws upon the technical

expertise of a highly trained workforce, and its startups almost immediately enter the US

market by forming an offshore office.

Another early location for software offshoring was Ireland. In the 1990s, both Irish

indigenous firms and multinationals rapidly increased their software-related activities in

Ireland. For the multinationals, Ireland was a convenient low-cost, English-speaking nation

that had strong European language skills. The multinationals adopted Ireland as an offshore

platform for Europe. In 2000, the total number of employees in the Irish software and

computer services firms, counting both indigenous and multinational firms, was 30,000

(Arora et. al. 2004). It is difficult to fully reconcile the various statistics as O’Riain (2004)

believes that in 1999 there were over 50,000 employed in the software industry. The larger

number may be misleading because O’Riain finds that the multinational firms that make up

approximately two-thirds of total software industry employment include among their

activities disk reproduction, packaging, language localization for Europe, and porting

(O’Riain 2004).

Ireland’s software industry includes two kinds of companies. There are some indigenous

firms that produce packaged products, although they have not been as successful as the

strongest Israeli firms. Ireland is also a packaging and localization platform for foreign,

particularly US, multinationals supplying the European market. Ireland has experienced

robust growth in its software industry but, like Israel, the global impact has been limited.

Recently, the Irish press has expressed concern that India might be a threat to the growth

of the employment in the Irish software industry (Weckler 2004).

Israel and Ireland were pioneers in entering the global software business without a

significant home market. Israel’s entry was at the high end of the industry, both in terms of

multinationals operating there and the local entrepreneurship. Because Israeli firms quickly

built strong business units in the United States, they are often treated as being the same as

US firms. In fact, the growth of a successful Israeli firm often occurs as much in the United

States as in Israel. In the Irish software industry, startups have produced software for the

world market but they typically remain small players, while multinationals located there are

fixed on one aspect of the global market, meeting the multinational’s localization needs for

Page 104

the European market. The software industry is a significant economic contributor to Ireland

even though it remains quite small in global terms.

Due to their small size and strong relationships with the rest of the developed world, the

Israeli and Irish software industries were successful without disrupting the software

industries in other nations. Wages in Israel and Ireland were slightly lower than the markets

they serviced, primarily the United States and Europe. Israel competed not on cost, but

instead on the high quality of its workforce. Ireland had a wage advantage and special

subsidies from the European Union but operated on only a small scale. What these two

countries showed was that a nation that did not have a large local market could

nevertheless perform software work at a distance from the final market if it had a skilled

workforce and access to good telecommunications infrastructure.

The next set of entrants could disrupt existing software industries. The largest and most

sophisticated of these national entrants was India. By the late 1990s, software

programming was no longer a skill that was highly concentrated in the developing nations.

People in low-income nations could afford an increasingly powerful personal computer and

had access to inexpensive, high-capacity data communications networks. Not surprisingly,

these countries could and did begin offering programming services in the global economy.

Today, it is possible to benefit from labor cost savings for programming services from a

large number of developing nations. Although the pattern is peppered with many

exceptions, there is a global division of labor emerging with India serving the Englishlanguage

market, Eastern Europe and Russia serving Western Europe, and China serving

Japan. Developing nations around the world have been eager to capture the wealth and jobs

associated with software offshoring.

From the inception of the computer industry in the 1950s, the United States was not only

the leading center for software but also defined the global software environment because of

its technology leadership, enormous market, and massive investment in IT R&D. Other

national markets were, for all intents and purposes, local markets having their own software

firms that were always under threat from being submerged by global firms. If local firms

wanted to expand significantly, then the US market was critical to their success. This was

something the Israeli firms understood from their inception. For this reason, the first two

nations to be discussed are the United States and its principal offshoring destination, India.

3.3 The United States

US-based companies continue to dominate the software and services industry. Of the

roughly $285 billion in total revenues of the global industry in 2004, only about $50 billion

was generated by non-US companies.1 2

History of the US Software and Software Services Industry

From the inception of the modern computer industry, the United States has been the

leader in both the hardware and software industries. The United States has also been the

source of many of the software standards such as Windows, Microsoft Office, and Unix,

1 This was calculated from the 2004 Global Software 500 ranking.

2 This accounts for only the traded software and software services. So, for example, if a firm writes

software internally for only Internal use, then this is not included because it is untraded. This is an

enormous category and is likely to be even greater than the amount traded. Chapter 4 examines

firms that have large internal software operations whose work is being offshored even though it is not


Page 105

providing US firms with an important first-mover advantage.3 Although today some might

dispute US leadership in hardware due to the growth of East Asian producers, few would

dispute US leadership in software and software services. Software and software services as

an independent business has been practiced in the United States for more than fifty years,

since the founding of the computer services firm Automatic Data Processing (ADP) in 1949.

Computer Usage Corporation (CUC), founded in 1955, was the first company formed

specifically to provide software development services to computer users. Its first project

was a program written for a customer to simulate the flow of oil. Computer Sciences

Corporation (CSC), founded in 1959, is now a $10 billion company. EDS, one of the most

important computer services firms, was founded in 1962. Since then, thousands of

companies that provide software and software services have been formed.

The growth of independent software and services firms was assisted by the decision by

IBM in 1969 to unbundle its application software and tools from its hardware. IBM did not

unbundle its operating systems from the hardware; the control system was included in

every product IBM sold for many years after 1969. By 1969, there were already

approximately 2,800 independent software product and services firms, and they had

combined revenue of $600 million (Steinmueller 1996). At the same time the software and

software services industry was emerging, large firms, especially in the financial and defense

sectors of the economy, were introducing computers into their operations and building

internal software competencies. For most firms, the building of internal IT expertise was

both a potential competitive advantage and a necessity because computers were becoming

key devices for managing the increasingly complex corporate operations they made

possible. By the end of the 1960s, the combination of government funding of engineering

and computer science research in the open university environment, early adoption by

sophisticated lead users, and the United States’ role as the largest economy and market in

the world meant that the United States gained what appeared to be an insurmountable lead

in the software arena.

The US software and services industry was affected by other developments as well.

Drops in prices of semiconductors and data storage, driven in part by Moore’s law, led to

continuous price-performance increases in computers. The big mainframe of the 1960s was

complemented by the arrival of the minicomputer in the 1970s and the personal computer

in the early 1980s. The PC drove the cost of a computer down to a level that permitted an

installed base of millions of computers, not the hundreds of computers of the 1950s or the

thousands of the 1960s. This growth in the installed base was accompanied by a huge

growth in the demand for packaged software for these computers. The early independent

software companies developed applications and later computer tools. Originally the

operating system software was provided by the hardware vendors (IBM, Digital Equipment,

and others), but in recent years, independent software houses have emerged that also

develop operating systems. (For a discussion, see Baldwin and Clark 2000).

The introduction of a commoditized personal computer in the 1980s and the spread of the

Internet in the mid-1990s led to the creation of many new US companies, not only

companies such as Netscape, providing software to facilitate the use of the Internet, but

also the service and shopping companies such as Yahoo! and The market

leaders, for instance, Google, Yahoo, Amazon, and eBay, weathered the dot-com stock

collapse beginning in 2000, and, in the process, they have transformed the way business is


3 It is possible that the relative strength of US firms might eventually be eroded by widespread

adoption of open source software.

Page 106

US firms benefited the most from the new business models and software that drove the

Internet, and these firms continue to be globally dominant. They were created from the

research and private sector capabilities that were uniquely resident in the United States

(Kenney 2003). The dot-com crash led to severe employment loss in the IT/software sector.

It was also a watershed event for the global software industry. During the height of the

boom, US companies could not find enough US workers and sought extra capacity from

overseas, especially Indian workers (both imported to work in the United States and

working in India). After the crash, the role of the Indian workers was more as a

replacement than a supplement to US workers. The Internet has also contributed to

creating a more global labor market, making it easier to access technical talent in any

location with good telecommunications linkages.

The Current Situation for US Companies in Software and Services

As mentioned earlier, US firms receive about 80% of the revenue available in software

and services. Of the top fifteen firms in this industry, only four – SAP (Germany), Hitachi

(Japan), CapGemini (France), and NTT (Japan) - are not from the United States, and these

firms occupy the bottom rungs of the top 15.

In 2004, US firms made up 16 of the top 20 packaged software firms when measured by

revenue. All of these firms have factories, development labs, and sales scattered across the

globe. But where is the employment? Of the approximately 595,000 workers in packaged

software, the United States employs 50 percent of the total global employment, while US

firms sell 84 percent of the packaged software purchased globally (McKinsey Global Institute


US firms have been remarkably successful. For example, Microsoft’s fiscal year 2005

profits of over $12 billion were comparable to the $12 billion in fiscal year 2005 revenue of

the entire Indian software and services export industry. Microsoft’s profit was also

approximately equal to the sales of the largest European firm, SAP. IBM’s software and

services revenues in 2004 were in excess of $61 billion. The point is that US packaged

software firms, by any measure, are still globally dominant. In terms of influence, the

importance of this dominance is even greater than simply sales; the United States is the

global hotspot for packaged software.

The McKinsey Global Institute (2005) reports that US jobs in the packaged software

industry are at risk of being offshored. McKinsey finds that 60 to 78 percent of the jobs at

risk are professional engineers and associated middle-level managers, that is, the heart of

the packaged software industry. Other occupational groups in the packaged software

industry have lesser but very significant numbers of jobs at risk. As shown in Table 2, it is

exactly in the more highly educated employment categories that US firms are recruiting

actively in India and, to a lesser degree, in China. Notice that the position announcements

are not confined to low-end college graduates but also include doctoral-level positions for

sophisticated development projects. The beginnings of this process can be seen in Table 1

which shows the number of employees the software and software services firms currently

have in India. In every case, these numbers are increasing at double-digit rates.

Page 107

Table 3-1: Indian Employment by Non-Indian Software and Software Services






in India




% in

India Locations

Oracle U.S. 6,900 (2004) 41,658 16.6 Bangalore,


Microsoft U.S. 1,250 (2004) 57,000 2.2 Bangalore,


SAP Germany 2,000 (2005) 38,802 5.2 Bangalore

IBM4 U.S. 23,000


369,277 6.2 Bangalore,





HP U.S. Yes 15,000


150,000 10 Bangalore

Veritas U.S. 900 (2004) 17,250 5.2 Pune

Adobe U.S. 500 (2005) 3,142 15.9 Delhi

Symantec U.S. 0 (2005) 5,300 0 n/a

EDS U.S. Yes 2,400 (2004) 117,000 2.1 Chennai,






France 0 4,088 0 n/a



France Yes 2,000 (2004) 59,324 3.4 Mumbai,



Bus Sys

Germany Yes 4,000 (2004) 36,000 11.1 Bangalore

Getronics Netherlands Yes n/a 28,000

Atos-Origin France Yes 750 (2004) 46,583 1.5 Mumbai

Tietoenator Finland Yes 120 (2005) 14,000 .9 Pune

* Hoover’s 2004

**Delhi includes Noida and Gurgaon which are suburbs in other states

All bolded firms include large non-software based employment

Source: Internet searches

4 This includes the 6,000 BPO employees when IBM acquired Daksh and also includes those working

for the domestic market.

Page 108

The United States leads in software services as well as in packaged software, and this

lead in services is also attributable in part to the early software development in the United

States. The United States is the largest single software services market in the world,

accounting for approximately 41 percent ($198.6 billion) of a total 2004 global market of

$484.3 billion (McKinsey Global Institute 2005). US vendors are the global leaders in the

global software services industry (11 of the Top 20 globally are headquartered in the United

States) with IBM Global Services at 2004 sales of $46 billion by far the largest. Software

services employment in the United States is approximately 1.7 million, of which 42 percent

are engineers (McKinsey Global Institute 2005). In other words, software services is a large

industry and the United States supplies about 32 percent of the total global workforce

(McKinsey Global Institute 2005: 158).

The occupational categories in IT services that McKinsey finds most amenable to

offshoring are software and hardware engineers and associated middle-level managers, of

which 47 to 56 percent could be offshored. Analysts working on software/IT architecture or

market research are similarly vulnerable (45 to 55 percent). It is in software services

where the most aggressive competition from Indian vendors is to be found, and where the

US leaders, such as IBM, Accenture, and Hewlett Packard, are rapidly increasing their

offshore and particularly Indian presence.

One can expect the number of available jobs, job tenure, and wages throughout the

software and service-related industries to be pressured by offshoring during the next

decade. This pressure will also be felt in the internal IT shops across all industries as

management considers options ranging from establishing offshore subsidiaries to

outsourcing the work to either a US firm operating abroad or an Indian firm. Routine

software production and services work appears to be increasingly susceptible to offshoring.

Conclusion on the United States

During the past five decades the dominance of the US industry has been a given. What is

changing is where the work will be undertaken. What has been an enormous export to the

world and a well-paid source of employment for technically well-trained Americans is now in

question as sufficiently well-trained individuals in much lower-wage nations are becoming

participants in the global economy and will be competing for those jobs. As Chapter 7 on

education discusses in greater detail, the US higher education system will have to address

the question of what their students should learn to prepare for these changes. At the

national level, there has been a dramatic underinvestment in engineering education and

research over the last two decades,5 and the recent decisions by the federal government to

reallocate research funds from universities to industry will further weaken engineering. This

is likely to contribute to an erosion of the cutting-edge research that makes the United

States a desirable place to undertake software innovation and development and which has

made the US high-technology industry a global leader.

3.4 India6

Software services have become India’s largest export, and the emergence of India as a

source of software service exports is attracting great attention in the developed world. India

has only recently attracted attention for its software service exports despite the fact that

5 For example, since 1970, U.S. federal spending in physical science research declined as a percentage

of Gross Domestic Product - an indicator as the rate of investment relative to

growth in the economy - from just under 1% of GPD to .5% in 2004.


6 This section draws heavily upon Dossani (forthcoming 2006).

Page 109

the industry has grown relatively steadily for three decades. Employment reached 697,000

(approximately 50% working for the domestic market) at the end of March 2005 (see Figure

2), a growth of 19.8 percent from the year earlier (Nasscom 2005). If the industry grows at

20 percent per year in 2005-2006, then the number of employees added in India would be

the equivalent of all the software workers in Ireland and Israel combined. India is emerging

as the single most important destination of software services offshoring.

Figure 3-2: Employment in the Indian Software Services and ITES-BPO Sector


As a large developing nation, India has many shortcomings including high rates of

poverty, corruption, and illiteracy; a substandard infrastructure; excess government

regulation; and various other problems typical of a poor nation. These obstacles are offset

by a number of strengths especially for software and services production. It has a long

history of producing capable mathematicians. It has a large population with adequate

English language capability. There is a large cadre of Indian managerial and technical

professionals working in North American and, to a lesser degree, in European hightechnology

occupations and organizations. For those who can afford it, India has a strong

and highly competitive K-12 educational system emphasizing science and mathematics.

Although India has a democratic socialist tradition with high levels of bureaucracy and overregulation,

it does have a market economy. These are all advantage that India has over

China in establishing a software services industry.

History of IT in India

The roots of India’s entry into the global IT industry can be traced to its initial highly

protectionist regulatory environment (Heeks 1996). As in many other nations, India’s

national policymakers focused on manufacturing. Protected from the global market and with

a domestic orientation, Indian hardware producers never became global competitors.

Because US firms established facilities in East Asia and homegrown Japanese, Korean, and

Taiwanese firms became subcontractors and later producers, the IT hardware industry

became concentrated in East Asia. Eventually, Taiwan emerged as the center for PC

assembly and India became largely irrelevant for electronics manufacturing (Dedrick and

Kraemer 1998).

The Indian software industry was established to serve the local market. Prior to the

decision in 1969 by IBM to unbundle its software from its hardware that spurred the growth

of an independent software industry, the only private Indian software firm was Tata

Consultancy Services (TCS) which had been established in 1968 to serve the in-house data-











42 70

FY00 FY01 FY02 FY03 FY04 FY05

IT Software and Services


Employee numbers ‘000s

Page 110

processing needs of the Tata Group. Using a Burroughs mainframe, TCS began offering

electronic data processing services to outside clients and also became Burroughs’ exclusive

India sales agent in 1970. India’s first exports occurred in 1974 when Burroughs,

recognizing the competence and cost advantage of the TCS personnel, asked TCS to install

its system software at the offices of its US customers (Ramadorai 2003 quoted in Dossani

2006). Aware of the profitability of providing such contracts, other domestic firms were

formed to offer similar services. Sending these programmers overseas to work on the

client’s premises became a common phenomenon, and was pejoratively known as bodyshopping.

Factors Contributing to Bangalore, India as a Principal Site of Offshoring

Bangalore is considered by most observers to be the hub of the Indian IT

industry. In fact, the United Nations Human Development Report has ranked

Bangalore as a global hub of technological innovation. The city of Bangalore is the

largest employer of software professionals in India, employing about 160,000

people in the technology sector of which IT services accounts for 100,000

employees, with the remainder in business process outsourcing and call centers.

The general context. Bangalore has had a number of advantageous events, some

historical and some recent, that have contributed to the rapid development of its

IT industry. These include the IT boom of the 1990s and the subsequent world

demand for IT products and services, the rapidly falling price of hardware, the

technological progress that enabled ever larger volumes of data to be copied onto

disks of the same physical size, the explosion of the Internet and the rapid

reduction in costs of sending data, the liberalization of the Indian economy in the

1990s, and the Y2K problem which came at the right time and showed that Indian

IT professionals could deliver.

High-tech center. From 1945, when Nehru became the prime minister, Bangalore

was considered to be the science city of India. The state of Karnataka, of which

Bangalore is the capital, is home to a large number of engineering colleges that

provide a steady supply of highly educated, skilled workers for the IT industry.

Bangalore also is the home to a number of large public and private sector

organizations that employ many specialized skilled personnel working in high

technology occupations. Though these organizations, such as the Indian Space

Research Organization (ISRO), the Hindustan Aeronauticals Limited (HAL), Bharat

Electronics, and Indian Telecom industries (ITI), are located in Bangalore and

were important for creating the technology-oriented environment, they have not

provided large numbers of technical personnel to the IT industry (D’Costa and

Sridharan 2003). Among Indians, Bangalore has a reputation as the technology

capital, though other cities such as Pune and Hyderabad are intent upon

challenging this perception.

Local government policy incentives. Policy liberalization was the tool used by the

Karnataka state government to create the environment that facilitated the growth

of the software industry. In the early 1980s, state officials made their first move

which was to establish a Software Export Processing Zone in Bangalore. Since

then the state government has acted to promote the industry’s growth in many

ways such as providing preferential treatment on land allocation, provisioning

electrical supply, and (until recently) providing a better infrastructure than most

other Indian cities. These incentives reduced the costs of setting up and operating

an IT company. (See Chapter 8 and later in this chapter for a discussion of the

national policy environment.)

Page 111

national policy environment.)

Telecommunications infrastructure. Texas Instruments had set up an office in

Bangalore in 1984 and petitioned the Indian Government for permission to lease a

64k line to be used for transferring data from India to the United States. The

Indian Government, both at the national and state levels, was worried what would

happen if they provided a 64k line to a foreign company, and it took three years

for them to approve the petition. Connectivity was through a local telephone

exchange in Bangalore that connected to the government-owned long distance

monopoly, BSNL. Service quality remained an issue.

By the early 1990s, both Western and Indian firms were demanding better

connectivity. Understanding the difficulties that business experienced when

interacting with the government bureaucracy, the government established the

Software Technology Parks of India (STPI). This organization was given

permission to provide last-mile connectivity, establish free trade zone status

parks, and generally facilitate the export software business. Bangalore was the

first STPI and has continued to be the most successful. STPI solved the

connectivity problem in Bangalore by installing a satellite dish antenna on its

property (more recently replaced by fiber optic cable). Despite the liberalization of

telecommunications in the late 1990s which led to drastically improved

telecommunications service and lower cost, there remains a role today for STPI

Bangalore that now serves at least 1300 companies.

Bangalore’s rise to prominence was due to a confluence of factors. It was

endowed with an excellent climate, a large pool of universities and governmental

research institutions, and a relatively robust physical infrastructure. In the 1980s,

it attracted US high-technology firms, such as Texas Instruments and Hewlett

Packard, to establish operations there by offering them international

telecommunications bandwidth. When these operations were successful, they had

a demonstration effect that attracted other multinationals. Significant political

support at the state government level ensured that the growth of the IT industry

was facilitated. These factors combined to make Bangalore the leading IT center in


The Indian scene changed in 1978 when IBM decided to withdraw from India, following

the passage of a law imposing joint ownership on all foreign subsidiaries. With IBM’s

withdrawal, the government formed and operated its own software firm, CMC. Though CMC

proved to be only moderately successful,7 it was in software and IT services that India

would become globally competitive. At the time, the Indian firms did little more than

recruiting, while an overseas intermediary secured the contract and the overseas client

decided on the work for the programmers who were sent to the client’s site. The initial focus

of this body shopping was on systems installation and maintenance. Later, the conversion of

clients’ existing applications software into (primarily) IBM-compatible versions began, but

this still operated on the basis of sending Indian workers to the client’s premises. By 1980,

the Indian industry earned $4 million in export revenue, shared between 21 firms, of which

TCS and a sister firm (Tata Infotech) accounted for 63 percent (Heeks 1996).

At the beginning of the 1980s, the Indian software industry was small, but it was earning

much needed foreign exchange. To encourage the growth of the IT industry and recognizing

India’s economic difficulties and foreign exchange shortages, Prime Minister Rajiv Gandhi’s

7 TCS purchased CMC in 2001.

Page 112

new government liberalized IT imports in 1984 through the New Computer Policy. Import

duties on hardware were reduced from 135 to 60 percent and on software from 100 to 60

percent. The software business was recognized as an industry, making it eligible for loans

from commercial banks. It was also delicensed, that is, permits were no longer needed to

enter the business. Delicensing was very significant because government licenses were

required in most of the Indian economy. Given that most sectors of the Indian economy

were off limits to new entrants, entrepreneurial energies were drawn toward deregulated

sectors. In addition to creating space for new Indian entrants, wholly-owned foreign firms

producing software for export were once more allowed, though on a licensed basis.

Electronics export processing zones were expanded to include software. TCS located in

the first of these that was opened in Mumbai in 1973. Rentals in the zones were set below

market levels and procedures to establish a business were simplified; power and water were

guaranteed. Most importantly, in 1985, all export revenue from these zones was exempted

from income tax (an exemption that is scheduled to end in 2007). These favorable policies

encouraged additional entrants.

These privileges, particularly the tax exemption, had a significant effect on the structure

of the Indian industry by making the domestic market comparatively less attractive; the

combination of a tax exemption and foreign currency earnings was irresistible. Thus the

Indian software industry was built on satisfying foreign demand for software services, not

products. India was on the verge of bankruptcy during the entire period, and the rapid

growth of the software industry and the fact that it was generating much needed foreign

reserves meant that the government developed a strong interest in encouraging its growth.

An important initiative to encourage growth was the creation in 1990 of the Software

Technology Parks of India system which was authorized to further simplify procedures and

enable exporters to import equipment against their export dollars without licensing or

customs tariffs.

These liberalizations were providentially timed because they coincided with an important

technical change in the software world, namely, the replacement of mainframes by

workstations that generally used the Unix operating system and C programming language

and were commonly linked together in a local area network (Dossani 2006). The adoption of

workstations as a work platform facilitated a gradual shift in the location of work from the

customer’s premises to remote production in India. Further, the adoption of these standards

generated work for Indian firms in converting clients’ installed applications into Unixcompatible

programs. The growth of this activity contributed to an increase in the number

of Indian firms from 35 to 700 by 1990 (Heeks 1996).

Even as the Unix-workstation standard became more prevalent, a number of

multinationals, including Texas Instruments, Hewlett Packard, and Digital Equipment

Corporation, opened wholly-owned subsidiaries in Bangalore to take advantage of low-cost,

high-quality Indian programmers to do various kinds of software-related development work.

As part of the recruitment package, the government agreed to supply them with then scarce

satellite bandwidth. Not much later, a few global banks with long-established Indian

operations, notably Citibank, also began producing custom software in India. By 1990,

Indian custom software developers were responsible for over 80 percent of all software


Despite obstacles, the Indian software industry grew and accumulated a number of

competencies. In 1991, the Indian government launched another wave of deregulation.

There were 700 firms, including several multinationals, operating in India at the time. Most

of these firms were small by international standards. The two Tata companies, TCS and Tata

Infotech, continued to dominate the industry capturing 48 percent of total revenue. Most

Page 113

firms usually had just one client and so were vulnerable to that client’s fortunes and

disposition. Two-thirds of the typical firm’s exports were to a single US client (Heeks 1996).

By the 1990s, the Indian government had become cognizant of the growing significance of

software exports and the need to encourage this one bright spot in a bleak industrial

climate. It now accepted that the industry required input from abroad. After the earlier

rollback of duties, by 1991 the duties on software had again risen to 110 percent. In 1993,

they were reduced to 85 percent; in 1994, they were further reduced to 20 percent for

applications software and 65 percent for systems software; and, in 1995, to 10 percent for

all software (Heeks 1996). Hardware duties ranged from 40 to 55 percent in 1995, but by

2000 had been lowered to 15 percent for finished goods such as computers, and eliminated

entirely for components.

The global software business was also changing as revenues in custom software overtook

product software. The custom software business was driven by the increasing size of

software programs that firms were using for their internal operations. Growing demand,

coupled with a shortage of US programmers, provided opportunities for the Indian IT

industry to offer its services. The Indian industry focused primarily on assisting in the

writing of the enormous software programs that were used inside large firms to control their

various business functions.

Indian sales efforts were handicapped by government regulations preventing them from

investing foreign exchange abroad. In the early 1990s, legislation was passed that allowed

firms to invest foreign exchange earned from exporting in order to establish offices

overseas. Previously, the Indian firms had only learned about their client’s needs from their

programming staffs on contract overseas, supplemented by occasional senior staff visits to

the United States (and occasionally to other high-wage countries) and client visits to India.

Operating a foreign office strengthened relationships with existing clients and provided

access to mid-sized firms. Some firms established dedicated centers at customers’ sites

(Dossani 2006). At that same time, the Indian government changed its regulations to allow

multinational firms to establish wholly-owned subsidiaries.

The Indian industry continued to evolve and find new software work. One important

opportunity was the Year 2000 (Y2K) problem that became a serious issue in 1998. In itself,

Y2K business was not so attractive as it was mostly unsophisticated work done at the

client’s site, but the Y2K business was important in other ways. It introduced additional

foreign companies to the abilities of Indian firms and programmers, thereby expanding the

Indian firm’s potential customer base and increasing awareness of India as a destination for

software work. Y2K prompted many firms to replace their legacy systems with standardized

software platforms such as Oracle and SAP. This meant that Indians could train on global

standard platforms and receive globally recognized certifications, raising client confidence.

The Current Status of the Software Industry in India

The Indian software and software services industry has experienced remarkable growth

over the last thirty years. The Indian software services industry is based on the use of

global software platforms and thus must purchase software licenses from foreign vendors.

There is little available data on Indian software imports, however. Heeks (1996) cites

Dataquest reports that in 1994-95 these imports were in excess of $96 million, and they

have almost certainly grown many times over as the Indian software industry came to

operate on standard platforms provided by US-based multinationals such as Computer

Associates, Microsoft, and Oracle. To provide some idea of the number of software-capable

people there are in India, according to Oracle (2005), India is the home to more than

220,000 members of its 3-million-strong online developer community, Oracle Technology


Page 114

Though this study concentrates on software offshoring, today not only software but also a

great range of services are being offshored to lower-wage-cost environments (Dossani and

Kenney 2003). Major software firms such as IBM, Microsoft, Oracle, SAP, and Veritas have

relocated work to India. Because of the way the data is reported, it is impossible to separate

the software work from the back office operations. However, as Table 1 indicates, a number

of these firms have large workforces in India. In the case of Oracle and Adobe,

approximately 16 percent of their global employment is now in India, and the number

continues to grow. Other major software firms, all of whose Indian facilities were

established far later than Adobe and Oracle, are also growing rapidly.

One myth about offshore facilities for multinationals is that their employment is limited to

relatively low-skilled programmers. In February 2005, Oracle was advertising for 199

positions in its two facilities in India. Approximately 30 percent of these positions were for

workers with Masters or Ph.D. degrees. Microsoft, newer in India, had a relatively less

highly skilled recruitment profile, but they, too, were recruiting highly credentialed workers

(See Table 2). Hiring such qualified employees, these Indian firms are likely to be capable of

innovation in the future.

Table 3-2: Educational Requirement Posted for Job Openings for Microsoft in India

and China by Highest Degree (Feb. 2005)


None Technical Bachelors Masters PhD Total

Beijing 2 0 0 1 0 3

Bangalore 2 0 13 5 0 20

Hyderabad 17 3 57 14 3 94

Source: Martin Kenney’s compilation from various corporate websites (2005)

Like the multinational software firms operating in India, the Indian-owned software

services firms are growing rapidly. They offer their services on outsourced software

projects. As Figure 3 indicates, the major firms are large. Headcount at TCS and at Wipro,

another major Indian firm, have already crossed 40,000 and are continuing to grow at 20-

25 percent per year (CAGR). The stock market believes in these companies and places

much higher values on them than on comparable US firms (Hira and Hira 2005).

Page 115

Figure 3-3: Total Headcount at Major Indian Software Firms by Year

Source: Heng 2005:7, Compiled by Martin Kenney from corporate sources

The prices offered by Indian firms place enormous pressure on management in

developed-nation firms to decrease costs so as to remain competitive. This resolves itself

into a single issue, namely, getting costs per employee down. The way to do this is to move

work to India or some other low-wage country. However, as Table 1 shows, many firms

such as EDS, IBM, SAP, and Cap Gemini have relatively low percentages of their workforce

located in India. From the of competitiveness perspective, this is no longer viable, and their

Indian headcount is expected to increase significantly over the next few years.

There is significant evidence that a movement to higher value-added activities is occurring

in both the Indian firms and the multinationals. In an Internet survey of the Top 86 U.S.

software firms as identified by Software Magazine and conducted in December 2004, 48

firms had R&D facilities in India, while 14 had facilities in China, and only three were

present in Russia (see Figure 4). There is also significant anecdotal evidence that US

software startups are establishing facilities in India to save money and increase their

headcount at low cost. (For further discussion, see Chapter 4.) This may have an indirect

impact on the future growth of US software employment.












2001-02 2002-03 2003-04 2004-05







Page 116

Figure 3-4: Number of R&D Operations in India, China, and Russia Operated by

Top 86 U.S. Software Firms

















Source: Internet searches by Martin Kenney

The Future of the Indian Industry

The Indian software industry is likely to grow in scale, scope, and value-added ability.

There is little reason to believe that offshoring as a process will end in the foreseeable

future, but it could slow down. The enormous investment by leading software multinationals

will expand the number of Indian project managers with global-class managerial skills. This,

plus the relocation of portions of startup firms to India, is likely to result in greater levels of

entrepreneurship and enable firms to sell their skills on the global market at global prices.

As a generalization, it is safe to say that this has not yet occurred, though I-flex, a former

Citibank custom software firm recently purchased by Oracle, is now selling proprietary

packages around the world (I-flex 2005). Other Indian independent firms may soon follow.

The offshoring of IT services and software for export will dominate the near future of the

Indian software industry. There are several possible trajectories. Custom projects could

become more complex and large, leading Indian software professionals to move from

programming into systems integration, systems specification and design. The average size

of projects Indian firms are undertaking has grown from 5 person-years in 1991 to 20 in

2003 (Krishnan 2003). As multinationals deepen their Indian operations, domain skills will

develop in India so that managed services are likely to become more important. This will

match global trends in the outsourcing of applications management and business processes.

Despite the fact that India’s software production for the US market exceeds that of any

other nation, it holds only a small share of the global market for all software value-added

work. The only part of the software value chain in which India has made substantial inroads

is in applications development where it has captured 16.4 percent of the world market. But

applications development is only approximately 5 percent of the entire global software

services market (see Table 3). This implies that there is much room for growth.

Page 117

Table 3-3: India’s Share in Various Sectors of the Software Services Industry





spending ($






revenues ($


Indian service

constituents by

percentage (%)




share of



Consulting 41.5 0.11 1.9 < 1



18.4 3.02 54.5 16.4

Managed services 124.9 1.94 35.0 1.6

System Integration:


Deployment and


91.7 0.37 6.7 < 1

System Integration:

Applications, tools and


62.4 0.10 1.8 < 1

IT education and


18.5 0 0 0

Total 357.6 5.54 100

Product software 200 1.66 <1

Source: Dossani from Nasscom 2004, p. 36 and 106. Indian figures are for 12 months ending March 2003. Indian

figures do not include product development and design of $ 0.56 bn and embedded software of $1.1 bn.

The Indian software and software services industries are booming. In 2004-2005, the

entire software and services industry grew at 18.5 percent and reached an all-time high of

$16.8 billion of which $4.8 billion was in the domestic market. The export earnings

increased at an annual rate of 30.4 percent from $11.2 billion in 2003-04 to $12.0 billion in

2004-05 (Nasscom 2005). All projections for 2005-06 indicate that it will be yet another

banner year.

A key issue for India is the future of applications development in the value chain.

Applications development may become commoditized just as systems maintenance has,

either due to automation or the development of products that are as good as custom

applications. Applications development has been losing global market share to consulting

and is slipping down the value chain. However, since information is a source of competitive

advantage, it is unlikely that customized application work will disappear altogether. In order

to grow, the Indian industry will have to shift to more complex activities by securing larger

projects, undertaking engineering services, integrating and managing services, or bidding

on projects that include transforming a client’s entire work process.

Increasing the value-added and IP components of Indian software services is difficult. For

example, Cognizant CEO Narayanan argued that India did not yet have the capability to

develop intellectual property, pointing out that R&D’s contribution to overall growth is

minuscule, and multinationals generally use their Indian R&D operations to upgrade existing

products, not develop new ones (Economist 2004). Sarnoff India head, Satyam Cherukuri,

Page 118

argues that India has two of the three requirements for innovation, technical skills and

access to capital, but lacks an “indigenous business model” (Economist 2004; D’Costa


Despite the assertion of many, it is plausible to argue that there is a significant

entrepreneurial movement emerging in India. It could be said that there have been two

clear waves of entrepreneurship already. The first was the establishment of firms such as

Infosys, HCL, and Hexaware who created body-shopping businesses that evolved into the

offshore programming model. Though entrepreneurial in genesis, they were pure labor-cost

arbitrageurs. More recently, that situation may be changing.

The second wave of entrepreneurs consists of a few startups that are producing their own

IP and marketing it globally. One of the most successful is I-Flex which was established by

Indian executives who spun out of Citicorp’s Indian software subsidiary. I-Flex developed a

banking software package that is now being used by more than 50 medium-sized banks

around the world. Today, there are only a few other examples, but given that an increasing

number of Indian managers and researchers are acquiring experience in the Indian R&D

laboratories operated by US firms, there is the potential for more of these startups.

Software offshoring to India is likely to grow not only through the continued growth of

indigenous Indian firms, but also because foreign software firms feel compelled to increase

their employment in India in product development and particularly in software services.

Including not only software and software services but also other services, Accenture hired

1,600 employees in May 2005 in India and has announced that it will be hiring 50,000 more

workers in India, China, and the Philippines in the next three years. IBM, which had 6,070

employees in India in 2002, saw the number rise to 24,150 in 2004. The company has a

target of raising this number to 38,196 in 2005, an addition of 14,000 employees in just

one year. CapGemini India plans to grow to 10,000 employees by 2007. Large

multinationals, such as IBM and CapGemini, are competing with Infosys, Wipro, and TCS for

offshore supremacy. As Indian companies move to global markets in their quest to expand

the offshore model, these large multinationals are moving to low-cost destinations, taking

the big Indian firms head on. The Indian firms are likely to face tough competition in the

near future.

The cost advantage in India may diminish as labor costs increase and the rupee

appreciates against the US dollar. However, interviews conducted by Rafiq Dossani and

Martin Kenney indicated that costs were generally increasing rapidly only for experienced

managers (15 to 20 percent per annum), while wages for beginning college graduates were

increasing more gradually (5 to 10 percent per annum) (Private communication, 2005).

Today, the cost of an Indian college graduate is $6,000-7,000 per year, while a US graduate

is in excess of $40,000 per year. In addition to wage costs increasing among the more

experienced managers, there is a generalized phenomenon of high turnover due to a

supply-constrained labor market. This turnover affects projects and may have a detrimental

effect on capability development both at the individual and corporate level.

There is much discussion of the high quality of the Indian IT labor force, but this may be

deceiving. NASSCOM indicates that only about 27 percent of the employees in the Indian IT

industry have an undergraduate or graduate degree in computer sciences or electrical

engineering. In spite of India having 247 universities and 11,549 colleges in 1997, only 7

percent of the student-age population attends a university (Nasscom 2005). India has 0.3

scientists and technicians per 1000 population, ranking 42 out of 62 countries as ranked by

the World Bank in 1998, below China at 1.3 (ranked 25th) and Ireland at 2.0 (ranked 20th).

This lack of highly educated workers may slow India’s advance into higher value-added

sectors of the software industry.

Page 119

Despite much improvement in the value-added per employee, India continues to trail the

United States in this regard. In India, revenue per employee in software services has risen

from $16,000 in 1990 to $33,000 in 2003. However, this is far behind the US average of

$142,000. This differential suggests that US workers are still more productive than those in

India, probably because a significant portion of the US revenue is in software products

where revenue per employee is much higher than in software services.

Improvement of the value-added per employee will require a continued upgrading of the

Indian workforce. The leading Indian software firms are investing in their workforces, but

there is only so much training an individual firm can undertake particularly in the high

turnover environment that characterizes the Indian labor market. Thus much of the

responsibility falls on the central government which is the main financier of tertiary

education. While India has greatly expanded the university system, problems with quality

appear to have deterred enrollment.

The interaction between university and industry is minimal. There are few academicindustrial

research partnerships as well as few consultancy assignments for faculty in

industry. On campus, little independent research is undertaken. Until recently, faculty (even

at the Indian Institutes of Technology (IITs)) have not been expected or funded to do

research. Only in 2005 did the Indian government appropriate $250 million to establish a

National Science Foundation. For example, at IIT Delhi, the value of sponsored research and

consultancy assignments in 1998 was only $4.5 million (Parthasarathi and Joseph 2002).

Faculty salaries are low, and NASSCOM concluded, “Over the years, there has been a

general decline in the quality of faculty in Indian universities” (Nasscom 2002). The average

number of citations over a five-year period for a faculty member at the Indian Institutes of

Technology is less than three. This compares with 45 per faculty member at MIT and 52 per

faculty member at Stanford University (Nasscom 2002). The country produces only 300

master’s degree graduates and 25 Ph.D.s in computer sciences each year, compared with

US numbers of 10,000 and 800, respectively. (For more information on the Indian

educational system, see Chapter 7.)

The Indian subsidiaries of multinationals are perhaps even more important than the

independent Indian firms. The reason is that the multinationals are more willing to

undertake high value-added activities such as software product development within their

own captive firm in India than they are to send the work to an Indian independent firm. It

is within these subsidiaries where the highest value-added activities, such as globally

directed research and development, take place. For at least the medium term, India should

be able to retain its position of primacy for software offshoring from the English-language

world. In the longer term, unless India makes an even greater effort to upgrade its

universities and the technical capabilities of their graduates, China may become an

important alternative destination.

3.5 China

China is one of the fastest growing economies in the world and is now the seventh largest

economy in the world in terms of gross domestic product and the second largest economy in

the world when the GDP is corrected by purchasing power parity. During the last two

decades, it has become a manufacturing powerhouse. In 2004, the United States had a

$162 billion trade deficit with China, the largest trade deficit with any single nation in US


8 These statistics can be found at

Page 120

China manufactures a broad range of goods, including IT products such as personal

computers, routers, monitors, cell phones, and handheld devices. The manufacturing of IT

products in China is growing more rapidly than China's overall industry. From 1990 to 1999,

the Chinese IT industry grew at a rate of 32 percent per annum (Dong 2004). From 2002 to

2003, sales increased at approximately 34 percent to reach $235 billion, and China became

the third largest IT equipment producer in the world (STAT-USA 2004). IT exports continue

to grow rapidly. For example, in the first seven months of 2003, China exported $80.6

billion, representing approximately 50 percent of its total production (China Venture Capital

Research Institute 2004). The strength of the Chinese IT hardware industry is shown by the

purchase in 2005 of IBM’s PC division by Lenovo. There is ample reason to believe that

China may soon become the largest IT product exporter in the world.

China and IT Software and Services

Much less is known about the Chinese software industry than is known about the Indian

software industry. The Chinese IT and software and service industries (ITSS) do not appear

to be having an important impact on the global economy, though as we shall discuss later in

the section on Japan, there are Chinese ITSS exports, and they are expanding rapidly but

from a far smaller base than in the case of India. According to the Chinese Software

Industry Association, there are 300,000 workers employed in over 6,000 firms, of which

approximately 160,000 are software professionals, approximately 25 per firm (Tschang and

Xue 2005, 133). According to the Ministry of Commerce, the revenues of the Chinese ITSS

industry increased from $7.17 billion in 2000 to $19.35 billion in 2003. During the same

period, software exports increased from $250 million to $2 billion in 2003 (China Software

Industry Association 2005). A recent report (Krishnadas 2005), notes that China's IT

services revenues are rising but are barely half of India's $12.7 billion. Growth is driven by

internal demand, and exports make up only 10 percent of total annual software service

revenues. The Chinese Software Industry Association indicates that 60 percent of Chinese

software exports in 2003 went to Japan and another 21 percent went to Southeast Asia to

nations using Chinese characters (Liu 2004).

Despite the impressive growth, the Chinese software export industry faces many

obstacles. It is extremely fragmented, and few firms are capable of undertaking large

projects (Krishnadas 2005). As of 2003, only six Chinese firms had received certification

through the Carnegie Mellon Software Engineering Institute's Capability Maturity Model

Integration, and most of these had not achieved CMMI Maturity Level 3 (China Venture

Capital Research Institute 2004). In contrast, all top 30 Indian software outsourcing firms

had already received CMMI Maturity Level 5 (Krishnadas 2005).

Given China's role as a manufacturing center for the global economy, many manufactured

products it exports contain embedded software. Programmers preparing embedded software

often need to work closely with designers and manufacturers, and there is some evidence

that a portion of this work may be relocated to China (Linden and Brown 2005).9 At the

moment, there are no estimates of the size of this market, but it could be significant.

The Chinese firms providing IT services to Western nations vary by size, but most are

relatively small. There are Chinese firms offering offshore software services for the US

market. At this point, the Indian firms dwarf them but, over the longer term, Chinese firms;;;

9 India also is developing a strong embedded software design capability, both at the multinational and

Indian firms.

Page 121

may provide competition to India in providing service to US businesses. A number of the

large Indian software offshoring firms have established subsidiaries in China for the purpose

of capturing business in China and servicing the Asia-Pacific market. Meanwhile, the

Chinese networking equipment firm, Huawei, employs nearly 800 engineers in India and has

announced plans to increase the number of engineers to 2,000 by 2006 (Press Trust of

India 2005). Although the outsourcing relationship between China and India remains tiny,

given the growth both nations are experiencing, it is possible that they could eventually

become significant.

In contrast to India where exporting is the goal of both the India independents and the

multinationals, much of the IT and software services growth in China is in the domestic

market. Multinationals have developed large operations bent on localizing their products and

software for the Chinese market. China is the largest developing country market in the

world, and, for certain products, it is rapidly becoming one of the overall largest markets in

the world. Many foreign goods must be localized to meet the special requirements of the

Chinese market. For this reason, China is becoming an increasingly important location for

R&D facilities in a wide variety of industries, including software and electronics (Zedtwitz

2004). In an effort to tap the Chinese market and utilize Chinese production prowess, a

number of US software firms as well as US, European, and particularly Taiwanese

electronics firms have established R&D facilities in China. The Shanghai area is an important

stronghold not only for computer machinery assembly, but also for semiconductor

manufacturing and, on a slower track, semiconductor design (Reuters 2005). Most of these

operations are geared to adapting products for the local market or doing production

engineering; however, some are developing global product mandates or are doing research

for the firm's global operations. One of the most celebrated of these is the Microsoft

research laboratory in Beijing which as of November 2004 employed approximately 170

scientists and planned to add 80 more (Heim 2003). According to Huang (2004), "more

than 70 technologies developed [there] are already used in Microsoft products, including

software for Windows operating systems and graphics packages for X-box video games.

More of the lab's latest software is slated for the next version of Windows due out in 2006."

Global software leaders such as SAP, Oracle, and Adobe are also establishing or expanding

their Chinese operations.

Given the general economic growth in China, the Chinese market for software is

expanding rapidly. Today, US packaged software firms are having some difficulty in the

Chinese market due to uncompensated software copying. Though China has joined the

World Trade Organization, it seems likely that these difficulties will continue. One possible

answer to the uncompensated copying is the current Chinese effort to move to open source

software (Marson 2005). Were this to come to fruition, it would dramatically decrease the

problem of uncompensated copying, but it would also have a significant impact on the

future prospects for growth of the US packaged software industry.

Conclusion on China

Given the past growth record, the apparent opportunities, and the importance given to

the software industry by the Chinese government, the Chinese software industry is likely to

continue growing more rapidly than the rest of the Chinese economy, and probably faster

than the software industry in the rest of the world. In the short term, the Chinese software

and software services exports are focused on Japan (see Section 3.6). In the long-term,

China could possibly emerge as a competitor to India in the general software export market.

More easily predicted is that China will become an enormous market. Given the relative

paucity of data, it is difficult to predict when China will become one of the largest software

markets in the world, but given the number of Internet and cell phone users, the rapid roll

out of broadband networks, and the gadget orientation of Chinese consumers, it might

Page 122

happen quickly. However, it seems unlikely that Chinese firms will be able to compete with

the major Indian IT and software services firms in the near-term.

3.6 Japan

The Japanese software and software services industry had sales of about $140 billion in

2004 and is the second largest single-country market in the world, accounting for 10.8

percent of the world’s IT industry. Further, IT and software services is the fastest growing

industry in Japan. In 2003, there were 5,482 information service companies employing

567,060 workers in Japan. Of these, technical positions included 240,096 system engineers,

114,479 programmers, and 7,398 researchers. The number of software engineers in all

industries is about 800,000. Thus, more than 40 percent of software engineers are working

in the information/service industry (JISA 2004a).

As Table 4 indicates, Japanese software imports were $2.9 billion in 2003.10 The type of

software imported is specific to the nation from which it was imported. The United States is

the largest source of software imports, and it overwhelmingly provides system and

applications software. In contrast, the imports to Japan from China and India are mainly

custom software. Japan imported $102 million worth of custom software from China and

$38 million from India. Japan also received $262 million in software services from China and

another $63 million from India (Umezawa 2005a).

Table 3-4: Japanese Software Imports in 2003 (US$ millions)


Applications Applications Custom




Outsourcing Total

U.S.A. 686 1,874 47 6 49 2,606

China 1 1 102 0 262 104

Ireland 0 45 0 0 0 45

India 2 0 38 0 63 40

Australia 0 0 26 0 26 26

Others 6 27 43 0 91 76

Total 695 1,947 6 2,901

The Japanese software industry differs significantly from that of the United States.

Outside of a few fields such as game software, Japanese software firms develop custom

software for the Japanese market. Few Japanese companies produce packaged software

products. For those firms that do prepare packages, their market is almost exclusively

domestic, and the entire industry, not counting game software, exports only $93 million.11

10 The true value of the imports is much larger because all major foreign software package firms have

subsidiaries in Japan. For example, Microsoft’s Asia-Pacific earnings in 2003 were $3.437 billion

(Microsoft 2004), and Japan is as large as the other markets combined so Microsoft alone probably

earned more than $1.5 billion in Japan.

11 For further information on Japan and offshoring, see Umezawa (2002; 2005a; 2005b).

Page 123

There is no authoritative data on offshore software development for the Japanese market.

The most reliable data comes from surveys conducted by the Japan Information Service

Association (JISA). According to Tsukazaki (2002), 19,000 foreign engineers were working

in Japan in 2001, and, during that year, 3,943 foreigners acquired the status of engineer. Of

these, 61.8 percent were estimated to be software engineers. In the JISA sample, Chinese

professionals were by far the largest group represented, followed by Koreans, and a

relatively small number of Indians.

The typical pattern in the past has been for Japanese firms to import Chinese or Indian

software engineers to work on projects in Japan. This has changed because the cost of

dispatching Chinese software engineers to Japan has increased to the point where it is no

longer attractive (Umezawa 2002). The more typical pattern in 2005 is that a Japanese

customer identifies a need for custom software and engages a Japanese software firm. The

Japanese software firm then contracts with a Japanese subsidiary of a Chinese firm to have

the work done either by Chinese programmers in Japan or, increasingly, by programmers

located in China.

Another mode of offshoring has emerged in which Japanese firms invest in China to form

a wholly-owned subsidiary or a joint venture with a Chinese firm. The most popular

locations for Japanese firms to operate subsidiaries in China are Dalian and Beijing. The

other mode of offshoring from Japan is for Western multinationals to move their

programming and back office functions from Japan to a lower-cost environment in China.

Dalian's software industry has grown (from a small base) at over 50 percent annually in

sales volume and reached $544 million in revenues in 2003 (Xiong 2004).

Japanese firms, such as Fujitsu, NEC, Sharp, and Sanyo have subsidiaries to produce

software in India. For example, through a publicly listed affiliate, Fujitsu (in 2005) employs

more than 2,000 workers in its four facilities in Pune. It is difficult to establish whether

these operations support the Japanese market or the English-language operations of

Japanese firms. Despite great effort on the part of Indian vendors, with only a few

exceptions, their sales in Japan have been small.

Impact on Japan of Offshoring

The impact on Japan of IT and software services offshoring is uncertain for two reasons.

First, the software services offshoring phenomenon is recent so that patterns are not yet

well established and data is scant. Second, Chinese vendors generally do not conduct

business directly with their Japanese customers so the role of the Chinese company is

somewhat obscured. There is typically an intermediary such as a Japanese software firm or

a US or European-based multinational, that holds the contract with the Japanese client; the

Chinese company is a subcontractor. The Japanese software firms have typically retained

the higher value-added activities in Japan.

Actual sales are relatively small. While sales of the Japanese information services industry

are $141.7 billion, the share offshored is $480 million or only 0.3 percent of the total sales

(Umezawa 2005a). This may underestimate the total amount of offshoring because many

multinationals operating in Japan have begun servicing the Japanese market from China or

India. For example, in 2004 the US consultant firm, BearingPoint, established a

development facility in Dalian with 60 employees, and it planned to increase employment to

1,000 “as soon as possible” (Thibideau 2004). Although Dalian has just begun to grow as a

software offshoring center, Western multinationals there already include IBM, General

Electric, Accenture, Dell, and SAP. Among the Japanese firms in Dalian are Sony,

Matsushita Telecom, Mitsubishi, Toshiba, Nokia, Omron, CSK, Alpine, Furuno Softech, FTS,

and Sino-Japan Engineering (Xinhua News Agency 2003). Despite this rapid growth, these

operations are still not significant enough to have much impact on the Japanese economy.

Page 124

The movement of Japanese software production to China is likely to continue. However,

there seems to be a division of labor emerging with Japan undertaking the higher-end

software development, while the Chinese subsidiaries and subcontractors undertake the

more mundane coding functions. This is borne out in the China Venture Capital Research

Institute (2004) assessment of the situation, “the export to Japan was mostly done in the

form of outsourcing, just like traditional manufacturing, what was subcontracted to our

country was only the development of lower-layer modules.”

Conclusion on Japan

The amount of work offshored to China is likely to increase due to the increasing pressure

to lower prices. Users are beginning to require that their Japanese vendors offer software

development prices that assume offshore development in China. Given the strong linkages

Japanese firms have with their customers, it is unlikely that Chinese firms will be able to

establish direct relationships with Japanese customers. However, Japanese firms will be

under pressure to reduce costs and thus will almost surely have to increase the scale of

their Chinese operations or form alliances with Chinese firms. This suggests that Japan will

not be immune to the pressure to offshore.

3.7 The European Union – Western and Eastern Europe

After the United States, the European Union as a whole is the second largest software

market in the world. In 2004, Germany accounted for 8.1 percent (15.4 billion Euros), and

the United Kingdom accounted for 7.1 percent (13.5 billion Euros) of the world software

market (Heng 2005). This is much smaller than the US share of 44.5 percent (96.6 billion

Euros). However, this statistic is somewhat misleading in that the Europe Union is not yet a

single market but is rather a loose confederation of markets with different customs and

languages. Europe has only one major software products firm (SAP). With the exception of

SAP, US firms are dominant in Europe’s packaged software market. Europe has no major

software service firms. The European market for IT services is divided by national language

differences. For example, Siemens Business Services is a leader in Germany, while Cap

Gemini is a leader in France. The giant US software service providers, such as IBM, Hewlett

Packard, and Accenture, play an important role in European markets where they usually

face firms that are only significant in their home nation (see Table 3-1). The greatest

European strengths are in software embedded in other products. Unfortunately, there are

few statistics publicly available to gauge the size of the embedded software market.

The European Union has been slower to embrace offshoring than the United States. The

United Kingdom was the first European country to do a sizable amount of offshoring, and it

is responsible for almost two-thirds of the IT and service jobs offshored from the European

Union (see Figure 3-5). Thirty percent of the jobs offshored come from Germany and

Benelux, and there is a noticeable increase in interest in offshoring in the German-speaking

nations recently. French, Italian, and Spanish firms have been more reluctant to send work


Page 125

Figure 3-5: European Offshore Services Spending by Region

Source: Parker 2004

Cost pressures are driving Continental European software and software services firms to

consider offshoring. According to a study by the consulting firm Roland Berger Strategy

Consultants (2004) in which executives at 93 major European firms were interviewed,

almost 40 percent of the firms have already relocated some services offshore, and 50

percent of all of the firms intended to offshore more activities. The firms already offshoring

gave the strongest indication of willingness to offshore new functions in the future. There is

anecdotal evidence that offshoring is increasing across Continental Europe. For example, in

2005, the Renault-Nissan alliance awarded IT services outsourcing contracts worth

approximately $600 million to two US firms, Hewlett Packard and Computer Sciences

Corporation, and the French firm Atos (Ovum 2005). With this outsourcing contract, much

of the work will be transferred to lower-cost environments since all three of these firms

have global operations.

Where the work is sent divides primarily along language lines. The United Kingdom sends

most of its work to India, while the Western European nations speaking languages other

than English are more likely to look to Eastern Europe. Due to the European Union’s

expansion into Eastern Europe, this is a natural near-shore location for the movement of

services. A significant percentage of Eastern Europeans speak a Western European language

such as German and, in the case of Estonia, Finnish. Hungary, Poland, and Romania are

seen as prime sites for this nearshoring work.

Although many studies predict that Eastern Europe and Russia will receive much of the

future offshoring work from Continental Europe, Figure 6 shows that, for Europe overall, the

most attractive location by a considerable margin will continue to be India. India wins on

price. For example, even though Romania is one of the lower-wage Eastern Europe

destinations, it still has wages that are higher than those in India. India also wins on

language and culture factors for the United Kingdom.








2004 2005 2006 2007 2008 2009


Million Euros








Page 126

Figure 3-6: European Offshore Service Spending by Recipient Location

Source: Parker 2004

Germany provides an interesting case study. A sample of 93 major German industrial

firms conducted by Horst Wildemann (2005) of the Technical University of Munich found

that, of the firms that have offshored, 29 percent sent their work to Eastern Europe, while

46 percent went to India and China. R&D and administrative functions were areas that

received considerable attention. This study also showed that the number of firms planning

to offshore service functions is increasing and predicts that Germany could lose 152,000

jobs per year, for a total of 758,000 jobs potentially lost in the next five years. A recent

study by A. T. Kearney (2004) predicts that by 2007 Germany will lose 130,000 jobs to

offshoring. Although German labor unions have resisted offshoring, the state government of

Bavaria is actively advising small and medium-sized enterprises on how to offshore their IT


Cost reduction is the prime driver of this offshoring for German firms. According to the

Roland Berger report (2004), the cost savings achieved through service offshoring were

typically in the range of 20 percent to 40 percent with an average of almost 30 percent,

which is similar to the results reported by Dossani and Kenney (2003) for US firms

offshoring to India. The wage differential between Germany and Eastern Europe is

significant. For example, in Germany, a systems engineer with a university degree and

three years experience earns between 57,000 and 63,000 Euros per year, while a

counterpart in Poland receives 15,000 to 18,000 Euros per year. Eastern Europe also has

the advantage in that there are no time zone differences to complicate communication.

Nevertheless, India and China are 50% less expensive than Eastern Europe which may be a

determining factor is some offshoring destination decisions.










2004 2005 2006 2007 2008 2009


Million Euros


North Africa

South Africa

! South Amer

SE Asia


Cent & East Eur

! Low-cost West Eur


Page 127

SAP and Globalization of the European Packaged Software Industry

SAP is the only major package software firm in Europe. Established by a group of

former IBM Germany managers and headquartered in Walldorf, Germany, SAP had

global sales of $9 billion in 2004. In keeping with its global reach, the company

has internationalized its sales, support, and development operations, and staffing

is growing in India in particular. The company employed approximately 32,000

people globally in 2004, including 13,500 workers in Germany, 5,000 in the United

States, and 1,500 in India. SAP's operations in Bangalore are not only at the low

end. For example, the Bangalore facility is developing software dealing

with international taxation which will be sold globally. However, with the

exception of SAP, it does not appear as though the European software industry

will significantly increase its strength in packaged software and thus, in packaged

software, it is unlikely that the employment offshored by the other software firms

will be large.

Case Study: Siemens Business Services and Globalization of European Software Services

Siemens Business Services (SBS) is a Siemens subsidiary that does software and

other outsourced production work on a global scale. It employs approximately

36,000 workers and has developed a strong business in helping companies to

implement SAP software. Its 2004, revenues of 4.8 billion Euros came from

Germany (48 percent), the rest of Europe (39 percent), the United States (8

percent), and the rest of the world (5 percent) (Siemens Business Services 2004).

Like many other large service firms, SBS has been globalizing its service delivery

operations and, in the process, downsizing its domestic workforce. Of the

company’s 36,100 global employees, only 15,100 are now located in Germany, and

4,000 are located in its rapidly growing Indian subsidiary.

SBS has developed a customer service strategy that uses a matrix of vertical

industry knowledge and sets of general competencies. One aspect of the matrix is

the industry expertise (vertical knowledge) concentrated in competency centers

that are scattered across different nations. For example, the paper and pulp

vertical center is located in Finland (Hallez 2004). The other part of the matrix

involves general activities, which are located in offshore sites in Canada, Ireland,

and Turkey to handle stabilized processes. India has two roles: it functions as a

back office operation for finance and accounting, and it does general software

programming and service and applications development for SAP programs. SBS

uses Russia for very labor intensive and repetitive back office and software

application development (Hallez 2004).

Conclusions on Europe

The European software industry and employment pattern is different from that of the

United States with much software production done in-house and embedded in physical

products. This does not prevent offshoring, and certainly many leading European industrial

firms are establishing offshore facilities to produce embedded software. Much of this

employment is subsumed under research and development and other activities such as

application-specific integrated circuit design that are not directly relevant to this report.

Continental European firms continue to lag the Anglophone nations in sending software

work across their borders. The Germanic and Nordic nations have only recently begun to

build offshore software and software service delivery capabilities, but firms with global

practices such as SAP, SBS, and others are moving rapidly to build their offshore capabilities

Page 128

in Eastern Europe, China, and India. The geography of European offshoring will be

somewhat different from that of the United States in that Nordic and German firms will use

Eastern Europe and Russia in addition to India. Those parts of Europe speaking a Romance

language as the predominant language have been slower to begin offshoring. But now their

major firms are sending work to Romania, Francophone Africa (particularly Morocco), and

Latin America, in addition to India. There is no reason to believe that the pressures to

offshore software-related work in these Romance-language countries will be substantially

different from in the Anglophone nations. Given that US-based multinationals with strong

global delivery capabilities, such as IBM, EDS, Hewlett Packard, and Accenture, are present

and competitive in all European markets, there are cost and delivery pressures on

companies throughout Europe to offshore, similar to the pressure on US firms. The only

possible mediating factors that will make the European and US situations different are union

and government opposition to offshoring in Europe. The most likely impact of this

opposition is a delay in offshoring, not a change in the final outcome.

3.8 Russia

In Russia, the largest state formed out of the former USSR, software was traditionally a

relatively neglected field outside the military. In the late 1980s, software comprised only 1.5

to 2 percent of the total outlays on computer systems, while the corresponding figure was

50 percent in the United States. At the time, there were reportedly few highly skilled

professionals among the country’s 500,000 programmers (Katkalo and Mowery 1996). In

the 1990s, Russia began a transition to a market economy, and many skilled software

engineers left the low-paid state enterprises, research institutions, and universities. They

either emigrated or moved to multinationals or Russian startups. The greatest international

product success of any of these companies was the computer game Tetris.

Large US and EU firms are active in Russia. Russia’s advantage is that, as a legacy of the

Soviet era, it has “more people working in R&D than any other country, and ranks third in

the world for per capita number of scientists and engineers” (American Chamber of

Commerce in Russia 2001). Russia’s investments in education mean that it has a large stock

of technically trained individuals. On the other hand, a recent Forrester report indicates

“while [the number of programmers in Russia] has increased during the past two to three

years, there is still less total development capacity than any of the large global system

integrators can provide in the United States or Europe alone (Hoppermann and Parker

2004)”. Even though Russia has trained technical personnel, the local software market is

small and undeveloped compared to those in Western Europe and the United States.

Russian firms have yet to play an important role in producing products or participating in

global software services.

Offshore software development in Russia represents a small fraction of the worldwide

offshoring headcount, although the number appears to be growing rapidly. Hawk and

McHenry (2005) estimated that the Russian offshoring software industry generated

revenues of between $200 and $450 million in 2003 and employed about 15,000 of the

70,000 programmers in Russia.12 The stock of potential programmers, that is, those with

some training in programming, may be as high as 200,000 and, in 2003 alone, there were

approximately 68,000 new graduates in electrical or telecommunications engineering,

computer science, mathematics, and physics (Hawk and McHenry 2005). Using the most

liberal definition of programmers, that is, college graduates from all disciplines who might

12 In 2001, the American Chamber of Commerce in Russia (ACCR 2001) estimated that there were

5,000 to 8,000 professional programmers in Russia doing $60 to $80 million per year.

Page 129

be capable of programming, it has been estimated that Russia could have graduated as

many as 225,000 in 2003 (Hawk and McHenry 2005).

Wages in Russia are low. In 2001, programmers with less than two years experience were

paid between $300 and $500 per month, while more experienced programmers earned

between $600 and $1,500 per month. Wage rates are increasing rapidly. Hawk and

McHenry (2005, 12) cited a 2003 survey that found that wages for development staff

ranged between $380 and $1,200 per month, with experienced managers receiving from

$700 to $1,900. Wages in Moscow were higher than these scales.

Contrary to the report cited, concerning the small number of high-skill programmers in

the Soviet era, Hawk and McHenry (2005) report that the skill levels of today’s Russian

programmers is quite high quality and they are considered to be good problem solvers. On

the other hand, project management skills are viewed as not so strong. Russia also presents

a difficult business environment. Experienced managers are in short supply, and few

Russian firms have secured certification from standards-setting bodies. Hawk and McHenry

(2005) state that only recently have Russian firms applied for certification, and only Luxoft

had reached CMMI Level 5 (Luxoft 2005). This may change over time as Russian firms

become more experienced and hire IT managers returning from abroad (American Chamber

of Commerce in Russia 2003). In addition, programmers with adequate English-language

capabilities are in short supply, bandwidth costs are higher than in most of the other

contracting nations, and the general legal environment in Russia is also quite uncertain.

The Structure of the Russian Software Industry

The independent Russian software industry consists of small firms. As of March 2005, the

largest firm, Luxoft, had over 1,000 employees (Luxoft 2005). There are a few other firms

in the 500 to 1,000 employee range. Despite this size limitation, Russian firms have won

business from important multinational customers, including Boeing, IBM, Dell, and Citibank

(Luxoft 2005; Hawk and McHenry 2005). A number of multinational corporations, including

Intel, Sun Microsystems, Motorola, Boeing, and Nortel, have opened R&D centers in Russia

to take advantage of the skills of Russian scientists and engineers. Intel is one of the firms

with the most ambitious plans for its Russian operations. In 2004, it purchased two Russian

technology companies, Elbrus and UniPro, increasing its total employment in Russia from

900 to 1,550 engineers and staff (Intel 2005). Sun Microsystems employs over 300 Russian

technologists in Moscow, St. Petersburg, and Novosibirsk (Nicholson 2004).

Conclusion on Russia

In terms of cost, quality, and volume, Russia is an attractive destination for offshored

work. There are a considerable number of capable, low-cost personnel available in Russia;

however, the stability of the business environment and the capabilities of management

preclude the type of massive growth seen in India or even China. The independent Russian

software firms are currently too small to tackle the largest and most sophisticated projects.

Russia’s strengths appear to be a number of technically sophisticated engineers capable of

doing cutting-edge research. Predicting Russia’s offshoring future is difficult because of

uncertainty regarding the continuing development of the system of higher education and

more general political and economic uncertainty.

3.9 Conclusions

Despite the changing geography of software and software services production, the most

important global relationship in software continues to be US firms providing software and

software services to the world. What is new is the perception by managers that capable

technical talent is available in developing countries, particularly India. For managers under

intense pressure to reduce costs, offshoring is now considered a normal response, and there

Page 130

is a growing infrastructure of lawyers, executive search organizations, and accountants in

place to facilitateit.

This chapter has reviewed the countries exporting work (e.g., United States, Western

Europe, and Japan) and their relationship with the countries that perform the work (e.g.,

India, China, Eastern Europe/Russia). A few conclusions can be drawn from this survey.

In the absence of major political or economic changes, the movement of software jobs

from developed nations to lower-wage environments will continue, perhaps at an increasing

pace, due to global markets, lower costs, and increased access to skilled labor.

Much of offshored IT work today is in lower skill areas, but this is changing. The change in

the nature of the work will require changes in the skills of the offshoring managers as well

as the employees who perform the offshored work. For the developed nations, it will be

critical to find ways to utilize this new resource of lower-cost IT workers to develop high

value products and services. This will require improvement in and the evolution of the

educational systems in both the developed and developing nations.

India has become the primary recipient of software and software services offshoring, and

this situation will continue for the foreseeable future. However, India is only the largest

beneficiary of the globalization of IT work. Any developing nation with properly trained

personnel, good telecommunications linkages, and the right cost structure can participate.

As a useful simplification, it can be said that India is the global center serving all

geographies, but that there are also regional divisions of labor emerging with Eastern

Europe and Russia tending to serve Western Europe, and China tending to serve the Asia

Pacific area, particularly Japan.

An emerging form of offshoring is the formation by multinational companies of in-house

laboratories located in lower-cost countries. This is one of the ways that the developing

countries can participate in more advanced research and development since traditional

offshoring does not lend itself as easily to advanced work.

3.10 Bibliography

A.T. Kearney Inc. 2004. The Future of Bay Area Jobs: The Impact of Offshoring and Key

Trends. San Francisco, CA.

American Chamber of Commerce in Russia (ACCR). 2001. White Paper on Offshore Software

Development in Russia (March).

Arora, A., Gambardella, A., and Torrisi, S . 2004. In the Footsteps of Silicon Valley? Indian

and Irish Software in the International Division of Labor. In T. Bresnahan and A.

Gambardella Eds. Building High-Tech Clusters. Cambridge University Press, Cambridge, UK,


Baldwin, C. and Clark, K. 2000. Design Rules: Volume One. The Power of Modularity The

MIT Press, Cambridge, MA.

Brown, C. and Linden, G. 2005. Offshoring in the Semiconductor Industry: A Historical

Perspective. In Susan M. Collins and Lael Brainard Eds. Brookings Trade Forum. Brookings

Institution Press.

Tarabusi, C. and. Vickery, G. 1998. Globalization in the Pharmaceutical Industry.

International Journal of Health Services 1, 67-105.

China Software Industry Association. 2005. China Software Export Achieved 7 Times Growth

in Five Years. Available at (Accessed March).

Page 131

China Venture Capital Research Institute. 2004. China Venture Capital Yearbook 2003.

China Venture Capital Association, Hong Kong.

D’Costa, A.P. 2003. Export Growth and Path-Dependence: The Locking-in of Innovations in

the Software Industry. In D’Costa and Sridharan, Eds. India in the Global Software

Industry: Innovation, Firm Strategies and Development. Palgrave Macmillan, New York, NY.

Dalian Software Park. 2005.

De Fontenay, C. and Carmel, E. 2004. Israel’s Silicon Wadi: the Forces Behind Cluster

Formation. In T. Bresnahan and A. Gambardella Eds. Building High-Tech Clusters Cambridge

University Press, Cambridge, UK, 40-77.

Dedrick, J. and. Kraemer, K.L.. 1998. Asia's Computer Challenge Threat or Opportunity for

the United States and the World? Oxford University Press, New York, NY.

Dong, Yuntin. 2004. The Chinese Experience. In P. Drysdale Ed. The New Economy in East

Asia and the Pacific. Routledge Curson, London UK, 130-139.

Dossani, R. To appear. Evolution of the IT Industry in India. In M. Hancock, W. Miller, and

H. Rowen Eds. Making IT. Stanford University Press, Stanford, CA.

Dossani, R. and Kenney, M. 2003.’Lift and Shift: Moving the Back Office to India.

Information Technology and International Development 1, 2, 21-37.

The Economist. 2004. India's Growing Strength in Innovation. (April 1).

Fujitsu India. 2005. Global Outsourcing Strategies. Available at (Accessed Sept.).

Hagel, J. 2005. Capturing the Real Value of Offshoring to Asia. Working Paper. Available at

Hallez, F. 2004. Global Sourcing and ICT Offshoring: The Approach of Siemens Business

Services. The EU-U.S. Joint Seminar on Offshoring of ICT Services and Related Services.

Brussels, Belgium (Dec.).

Hawk, S. and McHenry, W. 2005. The Maturation of the Russian Offshore Software Industry.

Journal of Information Technology for Development. To appear.

Heeks, R. 1996. India's Software Industry: State Policy, Liberalization and Industrial

Development. Sage Publications, New Delhi, India.

Heim, K. 2003. Microsoft Expands in China. San Jose Mercury News (Nov. 4). (Accessed


Heng, S. 2005. Software Houses: Changing from Product Vendors to Solution Providers.

Economics: Digital Economy and Structural Change. Deutsche Bank Research (April) #50.

Hira, R. and Hira, A. 2005. Outsourcing America. AMACOM.

Hoppermann, J. and Parker, A.. 2004. Debunking Russian Offshore Myths. Forrester

Research (June).

Huang, G.T. 2004. The World's Hottest Computer Lab. Technology Review (June). (Accessed March 2005).

I-flex. 2005. Home Page., (Accessed June).

Intel Inc. 2005. Intel Expands R&D Teams in Russia.

techresearch/research/rs08041.htm. (Accessed March).

Israeli Export and International Cooperation Institute. 2005. Israel’s Software Industry.

(Accessed June).

Page 132

JISA (Japan IT Services Association). 2004a. The Export and Import of Software and

Foreign Software Engineers in 2004. Tokyo, Japan.

JISA (Japan IT Services Association). 2004b. IT Services in Japan. Tokyo, Japan.

Katkalo, V. and Mowery, D.C. 1996. Institutional Structure and Innovation in the Emerging

Russian Software Industry. In D. C. Mowery Ed. The International Computer Software

Industry. Oxford University Press, New York, NY. 240-271.

Kenney, M. 2003. The Growth and Development of the Internet in the United States. B.

Kogut Ed. The Global Internet Economy MIT Press, Cambridge, MA. 69-108.

Krishnadas, K.C. 2005. Fragmented China Software Sector No Match For India, Report

Finds. EE Times (Feb. 21).


Krishnan, K.A. (Vice President, Tata Consultancy Services). 2003. Personal Interview by

Rafiq Dossani.

Liu, X. 2004. Technology Policy, Human Resources, and the Chinese Software Industry.

Conference on Strategies for Building Software Industries in Developing Countries. Hawaii.

(May). Hawaii_SW_Conference/Liu%20Paper.pdf.

Luxoft Inc. 2005. (Accessed March).

Marson, I. 2005. China's Linux Market Grows. ZDNet UK (March 15) 1.,39020390,39191343,00.htm.

McKinsey Global Institute. 2005. The Emerging Global Labor Market: Part I – The Demand

for Offshore Talent in Services (June).

Microsoft Inc.

Nasscom. 2005. Indian IT-ITES – FY05 Results and FY06 Forecast (June).

Nicholson, A. 2004. Intel Adds Russian Research Centers. St. Petersburg Times (May 28).

Oracle, Inc. 2005. Oracle India Factsheet.

factsheet.html. (Accessed June).

Ó’Riain, S. 2004. The Irish Software Agglomeration: Technology Driven Commodity Chains,

Global Regions and the Developmental State. International Journal of Urban and Regional

Research 28, 3, 642-663.

Ovum Inc. 2005. Renault Gives the French Outsourcing Market a Kick-Start. (Accessed July).

Parker, A. 2004. Mapping Europe’s Offshore Spending Impact. Forrester Research Inc.


Parthasarathi, A. and Joseph, K. 2002. Limits to Innovation in India’s ICT Sector. Science,

Technology and Society 7, 1, pp. 13-50.

Press Trust of India. 2005. Huawei India Gets New COO, Plans R&D Campus in Bangalore.

(May).,00030001.htm. (Accessed Sept.).

Reuters. 2005. China Emerging as Chip Design Center. (Aug.). (Accessed Sept.).

Roland Berger Strategy Consultants. 2004. Service Offshoring (June).

Shah, R. 2005. Ketera India Case Study. Conference on the Globalization of Services.

Stanford University (June).

Page 133

Siemens Business Services. 2005. Corporate Website. (Accessed March).

Sigurdson, J. 2004. The Emergence of the People’s Republic of China: Challenges and

Opportunities for Latin America and Asia. LAEBA Annual Conference. Beijing, China (Dec.). (Accessed March).

Singh, J. 2003. Country Analysis: Mexico.

Country-analysis-mexico.html (Accessed June 2005).

STAT-USA. 2004. A Year-End 2003 Summary of China’s Electronics Information Industry

Market (March).

(Accessed March 2005).

Steinmueller, W.E.. 1996. The U.S. Software Industry: An Analysis and Interpretative

History. In D. C. Mowery Ed. The International Computer Software Industry. Oxford

University Press, New York NY, 25-52.

Thibideau, P. 2004. BearingPoint Opens Second Development Facility in China.

ComputerWorld (July 6)


Tschang, T. and Lan Xue, L. 2005. The Chinese Software Industry. In A. Arora and A.

Gambardella Eds. From Underdogs To Tigers: The Rise And Growth Of The Software

Industry In Brazil, China, India, Ireland, And Israel. Oxford University Press, New York, NY,


Tsukazaki, Y. 2002. Requirement and Situation of Acceptance of Foreign Workers with

Special Reference to IT Engineers. in The Report of the Study Committee on Employment

Management of Foreign Workers in the Field of Specialists and Technology (March).

Umezawa, T. 2002. The Situation of Foreign Software Engineers- The Division of Labor of he

Software Industry and International Immigration. The Report of the Study Committee on

Employment Management of Foreign Workers in the Field of Specialists and Technology


Umezawa, T. 2005a. Report on Foreign Software Engineers in Japan. Waseda University


Umezawa, T. 2005b. Case Study on the Chinese Software Iindustry in Beijing. The Division

of Labor of Software Industry (April).

UNCTAD. 2004. World Investment Report 2004: The Shift Towards Services. United Nations,

New York, NY.

Weckler, A. 2004. The India Jobs Timebomb. Sunday Business Post (Jan. 4). (Accessed June


Wildemann, H. 2005. Unternehmensstandort Deutschland Wege zu einer

wettbewerbsfähigen Wertschöpfungsgestaltung. TCW Transfer-Centrum, Germany.

Xinhua News Agency. 2003. IBM Launches R&D Center in Dalian. (Feb. 9)

Xiong, S. 2004. Dalian To Build Software Industry Corridor. Xinhua Newsnet (July 4)

Dalian%20to%20build%20software%20industry%20corridor.phtml (Accessed June 2005).

Zedtwitz, M. von. 2004. Managing Foreign R&D Laboratories in China. R&D Management 34,

4, 439-452.

Page 134 *O. Berry and M. Kenney took a leadership position in the writing of Chapter 4.

Authors: Alok Aggarwal, Orna Berry*, Martin Kenney*, Stefanie Ann Lenway, and Valerie


Chapter 4: Corporate Strategies for Software


4.1 Introduction

The previous chapter examined the provision of software production and services for the

global market from a national perspective. This chapter examines the same issue from the

perspective of the firms involved. Our interest is in how firms use low-wage environments to

undertake software work for their global operations. The interest here is not in the

relocation of work from companies in a high-cost nation such as Germany or Japan to firms

in another high-cost nation such as the United States or the United Kingdom even though

this type of relocation is common. This chapter also omits work sent to Canada which does

have somewhat lower wage rates than the United States and is the beneficiary of what

some have termed near-shoring from the United States. Also excluded from this

presentation are the operations of multinational or domestic firms that service the local

economy of a low-wage nation. In most cases, these are relatively small operations except

in the case of China whose domestic consumption of software is increasing rapidly.

This chapter does not debate the reasons firms offshore to nations with significantly lower

wages; it simply accepts the fact that the wages are significantly lower. (An analysis of the

various reasons firms send work offshore is presented in Chapter 1.) In addition to reducing

costs, a company’s decision to offshore is often dependent upon two dimensions. The first

dimension involves its strategic decision regarding the kind of human capital that it would

be able to access when it goes offshore. Put differently, it is an uncontroversial observation

that if the nation where work is sent did not offer the proper skill set in its workers, firms

would not relocate work to that environment. The second dimension is cost. It is

uncontroversial to state that, given the right skill set,1 a sufficiently low cost of labor, and

work that can be done remotely, firms will find it attractive to locate to that environment

even if there is no market in that locale.2

A decision to offshore software work may come about in a variety of ways and may take a

variety of forms. The decision to offshore work has traditionally been made by the manager

with responsibility for a project, including profit and loss responsibility. In cases where the

contract involves strategic operations, critical company proprietary information, or very

large budgets, for example, the decision is often made at a higher level in the organization,

sometimes as high as the CIO, CFO, or CEO. In certain cases, the real reason for offshoring

might be simply that competitors have already done it or the board of directors is

demanding an offshoring initiative to save money. Rarely are the answers so simple, but

there are numerous anecdotes about how an executive team in the United States will

demand that the overseas operation achieve a certain headcount reduction by a clearly

1 This chapter focuses on the availability of technical personnel as the attraction for offshoring. It is

important to add that capable managers are also extremely important. As Parthasarathy (2005)

points out, the executive management team is critical for the success of an offshore subsidiary.

2 For example, Nike produces athletic shoes in a large number of nations where there are few, if any,

customers for its shoes.

Page 135

unrealistic date. The responsible executives will achieve the headcount goal regardless of

the economic justification. In other words, rationales for action vary. Moreover, similar

firms often have different recipes for using offshore resources. One basic decision a

company that has decided to offshore must make is whether to undertake the work in its

own offshore premises or outsource it.

This chapter considers five kinds of firms that are involved with software production or

software services that are provided by one or more developing nations.3

1. Packaged software firms headquartered in developed nations that make and sell

software as a product, for example, Adobe, Microsoft, and Oracle.

2. Software service vendors headquartered in developed nations. These companies may

also provide packaged software, though not all of them do so. Examples include IBM,

Accenture, and EDS.

3. Internal software operations in firms headquartered in developed nations that have

software operations but are not part of the software industry. This encompasses all the

companies producing non-IT goods and services. (The group is eclectic and enormous.

The importance of this category is that software is now at the heart of value creation in

nearly every firm. This is true of financial firms, such as Citibank and HSBC, and

manufacturing firms, such as General Motors and Siemens. Each of these companies has

a large staff writing software. To illustrate, it is estimated that by 2010, 40 percent of

the value of an automobile will be in its electronics of which embedded software will

become an increasingly important component. In 2002, it was estimated that the typical

luxury automobile had 105 microprocessors. up from 70 in 1998 (Tsai 2004).)

4. Software-intensive, high-technology startups based in developed nations. (This

category, though small in numbers of jobs, is important because these firms provide

many of the jobs of the future. For these firms, frequently there is no job displacement

at all. Rather it is the location of the future employment growth that is in question.)

5. Offshore IT service providers headquartered in developing nations that provide services

for firms in the developed nations. As was discussed in Chapter 3, firms providing

software services have emerged in a number of countries, though the largest by far are

located in India.

This classification of firms is only heuristic. The global Fortune 1000 firms have

complicated webs of relationships which might include newly built facilities, facilities they

acquired, contractors from developed nations (e.g., IBM and Accenture), and contractors

from developing nations (e.g., Infosys (India), I.T. United (China), and Softtek (Mexico).

Some product firms outsource certain activities to contract R&D firms and even form joint

ventures. There are also intermediate solutions such as the build-operate-transfer option,

which lies in between building one’s own facility and outsourcing. The tasks being

3 We do not use the term captive in this chapter, even though it is used elsewhere in this report. In

keeping with the literature on international business, we use the term subsidiary. It is more accurate

and does not suffer from the bias reflected in the term captive. The categories are not divided on the

basis of which firms are subsidiaries because #1, #2, #3, and #4 are all subsidiaries. Only the

developing nation firms’ operations are not subsidiaries. The categories developed in this chapter are

for the purposes of understanding the impact of globalization on software professionals and thus they

may not be useful for other purposes. For example, if one was merely interested in globalization, it

might be that the packaged software firms should be combined with software services firms. Or,

alternatively, software and software services might be combined. The separation of small start-ups

from large software and software services firms is justified only because of the importance they have

for the high-technology economy. For other purposes, this separation might not be proper.

Page 136

undertaken vary widely and include activities such as low-level software support, product

testing, product development, and research and development. The options and

permutations are numerous, and the case studies in this section are merely overviews, thus

they cannot do full justice to the breadth and scope of the software and software service

operations of these firms.

For each category, the chapter gives a general discussion of the outsourcing issues faced

by a specific kind of firm, followed by several case studies to illustrate the types of

operations the firm carries out in developing nations and why those particular countries

were chosen. These case studies are intended to be illustrative but not exhaustive. The

particular cases were selected in order to provide a balance across sizes of companies and

do not constitute a random selection upon which generalizations should be based.

4.2 Offshoring Firms

Large, Established, Developed-Nation Software Firms

Because of the somewhat different dynamics of the packaged software firms and the

software services providers, we discuss them separately despite the fact that there is

significant overlap between these two categories of firms. For example, Oracle and

Cadence, which are usually considered packaged software providers, have large consulting

arms to assist with the installation and operation of their software. IBM sold $15 billion in

software in 2004, yet it is today more of a software services firm (with revenues of $46

billion in its global services unit). Accenture is a massive consulting firm that provides a

variety of services, including software services. Thus the line between the two categories is

somewhat difficult to draw, but it is nevertheless a worthwhile distinction because a pure

packaged software firm such as Adobe or Microsoft hires programmers almost exclusively,

while a firm such as SAP, IBM, or Oracle also hires consultants and analysts who are not

necessarily working on products but are providing services.

Packaged Software Firms

The packaged software firms are what most people think of when they think of software.

As a general rule, the largest and most successful packaged software firms in the world are

headquartered in the United States (the notable exception is SAP in Germany). Successful

packaged software firms can be very profitable because they only need to write an

application a single time (although perhaps in several variations) and then reap their

revenues from the sales of many copies. One reason for establishing offshore facilities is to

localize the package for particular language groups. For example, Ireland has a large

industry that specializes in localizing products from US software firms for the European

markets (O’Riain 2004). Localization work characterizes a significant portion of the work by

the R&D laboratories of packaged software firms in various nations. This type of work,

though important for the global economy, is not of particular interest here.

There are, of course, other motivations for package software firms to locate in developing

countries. The most frequently given reason is access to the talented labor force working in

these lower-cost locations.4 One important motivation behind offshoring for these package

4 The decision to move to a location for lower cost is a complex one. Lower cost includes not only

wages but also the lower cost of benefits including health care. It also includes issues such as

reduced concerns about discipline problems, substance abuse in the workplace, and governmental

regulations concerning harassment, racial policies, etc. that are part of the protections commonly

expected in the developed nations. This chapter does not place a judgement upon these policies.

Quite naturally, in each nation, there are different regulations and standards that channel business

activities and create various costs and benefits.

Page 137

software producers is that their packages are constantly increasing in size and complexity,

driving the cost of writing the software, testing, and debugging it ever higher. Whichever of

these causes is the most significant, what is certain is that nearly all major packaged

software firms are establishing offshore facilities in lower-cost environments, ranging from

Eastern Europe and Russia to India and China. In the following case studies of Adobe and

SAP, we examine this new geography of the software industry.


On account of its Acrobat program, Adobe has a wide global footprint. Its software has

applications in digital imaging, design, and document technologies. The firm does its

product development in the United States, Canada, Germany, Japan, and India. In India,

Adobe has its largest physical office space outside of the United States, and the Indian

operation is growing more rapidly than any other location.

In 1997, Adobe established a sales office in New Delhi, India, to market its products. In

1998, it established an R&D center in New Delhi (Noida) to utilize the low-cost R&D talent

available in the country. By 2005, Adobe had 3,800 employees worldwide and

approximately 500 (13 percent) were located in India. Adobe has invested $10 million in

India but plans to increase that to $50 million over the next two years as the R&D center

grows. Adobe was perhaps the first international software company to develop a full-fledged

product in India, Page Maker 7 (Rediff 2005). The Indian center has filed 25 patents in the

last four years, an indication of the sophistication of the work it is undertaking.

In 2005, Adobe acquired Macromedia, another Silicon Valley firm. Rather than

consolidate Macromedia’s Bangalore research operation into its own research operation in

New Delhi, it is retaining and expanding the Bangalore facility which in April 2005 had 150

workers and was expected to grow to 250 by year-end 2005 (Verma 2005).

Adobe’s Indian R&D center works on Adobe Acrobat desktop applications and serverbased

products as well as products related to digital imaging and video. It develops

components for almost the entire range of Adobe’s product line. Products it has

worked on include PageMaker, FrameMaker, Postscript, Photoshop Album Starter

Edition, and the Acrobat Reader on Unix and alternate platforms.

From the managerial perspective, the Indian operation is becoming increasingly

integrated into Adobe, as is evidenced by the fact that Naresh Gupta, who managed the

Indian operation since its inception, is being relocated to the San Jose headquarters where

he will join the executive management team (Rediff 2005).

There can be little doubt that India has become Adobe’s low-cost development center. To

date, Adobe has not established development centers in other low-cost countries. In 2002,

there were articles in the press stating that Adobe might abandon sales in China because of

concerns over software piracy; this was quickly denied by Adobe spokespersons, but the

company has not moved to open a development center there (Sim 2002). There are

indications that Adobe’s Indian operations will continue to grow at least through 2007. If the

last four years are any indication, the percentage of the company’s employees located in

India is likely to increase (given that the total global headcount is growing slowly).


SAP, established in Walldorf, Germany, is one of the world’s largest software vendors with

operations throughout the world. SAP Laboratories is its R&D arm and has units in nine

locations worldwide: Walldorf, Germany; Palo Alto, California, United States; Bangalore,

India; Tokyo, Japan; Sophia Antipolis, France; Sofia, Bulgaria; Montreal, Canada; Tel Aviv,

Israel; Shanghai, China; and (most recently, in 2005, with about 50 employees) Budapest,

Hungary. The role of SAP Labs is to distribute global development efforts, enable SAP to

access the world's best IT experts, support local and global markets, develop first-class

Page 138

solutions, and drive innovation and competitive advantage for SAP, its customers, and


Most of the laboratories are relatively small. For example, SAP’s fourth largest lab

worldwide is in Israel and it employs 500 people (it also has another, smaller R&D operation

in Israel located within a firm that it acquired), while Sofia, Bulgaria employs only 200

programmers (SAP 2005). Each laboratory has its own specialties. The Shanghai laboratory

has focused on localization work, but is slated to grow to 1,500 engineers by 2009 and is

expected eventually to do more than just localization (People’s Daily Online 2004). In China,

SAP is cooperating with the Chinese Linux supplier Red Flag Software to develop corporate

applications for Linux (Bishop 2005).

The India SAP Laboratory was established in 1996 and has grown to be the largest

lab outside Germany with 2,000 employees today, and employment expected to

reach 3,000 by 2006. According to SAP AG executive board member Shai Agassi,

"Indian developers had contributed substantially to the global success of the

NetWeaver, the first appli-structure platform for enterprises across the verticals. The

SAP Labs India team is one of our most important development teams for NetWeaver

worldwide" (Indo-Asian News Service 2005). Even though the Indian operations for

the global economy are categorized as research and development, many of the

employees are in the services and consulting operations (Barlas 2004).

SAP has a global R&D and operations strategy with its various laboratories

specializing in different areas of software development. The plans for employment

growth for both India and China are aggressive. If current plans are realized, India

and China will become even greater portions of SAP’s total global headcount. Given

their economic growth, these countries will also become sizable markets for SAP.


The large package software firms are building increasingly global operations. In many

cases, their offshore operations are for localization work for the domestic market. However,

particularly in the case of India, but also in Russia, the work is for their worldwide software

packages. Locating in low-wage countries enables these firms to have access to lower-cost

programmers, many of whom are comparable in skill levels to the workers in the developed

nations. This is not the only benefit. Having operations in other time zones can speed up

production by facilitating round-the-clock production. These opportunities are encouraging

the rapid expansion of employment by major packaged software firms in India and other

lower-cost nations.

Offshoring will have a complicated effect on the packaged software firms and developed

nations. First, it might, and likely will, put employment pressure on software firms to

decrease employment in the developed nations. Alternatively, the lower cost and faster

production could allow the development of new features in old software and could contribute

to the production of lower-priced software products, thereby increasing usage that could

result in higher revenues and greater hiring. If the trends as described in these case

studies continue for the packaged software firms, elements of both of these scenarios may


Software Services Providers

Software service firms have been among the fastest growing firms in the IT sector, and in

general they are far larger than the packaged software firms. This section confines

discussion to the software service activities of these firms, but it is important to remember

that firms coming from the software service side (such as IBM or Hewlett Packard) and from

the service side (such as Accenture) are converging. In the case of IBM, this has been

achieved both through hiring and its recent acquisition of the Indian service firm Daksh

Page 139

(with its approximately 6,000 employees). For service providers, software and various other

software-based services (i.e., anything done on a computer) may be converging. The

software services firms are basically in what might be called a headcount business; they

grow by hiring more workers. Thus they tend to have more employees than most of the

packaged software firms.


Established in 1911, IBM has been the global leader in computer hardware and software

products and services. In this section, we focus on three different IBM activities, namely,

software products, software services, and research and development. It is important to

understand IBM’s scope and scale. In 2004, it had annual revenue of approximately $96

billion. Global headcount at the end of 2004 was expected to be more than 330,000

employees, excluding employees gained from acquisitions and strategic outsourcing

contracts (IBM 2004). The company’s geography of revenue growth is shifting dramatically.

In Brazil, China, India, and Russia, IBM’s annual revenue growth from 2003 to 2004 was 25

percent (though from a small base), while growth in the developed nations was on the order

of 4 percent. Between 2002 and 2004, IBM increased its workforce in these four nations by

30 percent (Palmisano 2005).

As of 2005, IBM’s Software Group had revenues of $15 billion and contributed one third of

IBM's profit. In the Asia-Pacific region, this group employed 5,000 people, including sales

and marketing. In India and China, IBM’s software development laboratories employed

1,500 in each country (Smith 2005).5 Richard Smith, the vice president of the Asia-Pacific

region for IBM Software, stated that “the Chinese market is internally focused. In India, a

lot of the software development activity is mixed - it is focused internally as well as on

exports.” As an example of the contributions of offshore centers, “a significant chunk of the

code for its AIX version of the Unix operating system was developed in India.” (Smith 2005)

IBM is already well advanced in using global software development teams. Hayward

(1997) described a global application development team it created that uses two

shifts. The first one is a small group of 25 people in Seattle that would set a daily

work specification for a particular application and assign it to offshore teams of 31

programmers each in India, China, Latvia, and Belarus (a former Soviet republic).

The offshore team in each location would write code to those specifications during

their daytime work hours. The code would then be sent back to Seattle where it

would be reviewed and tested. In principle, this process should not only lower labor

cost but also accelerate production.6

Software development at IBM is now a global process with the offshore low-cost nations

growing rapidly to meet increasing demand. In the illustration by Hayward, the Seattle

team was clearly the dominant team. However, given the increasing capabilities in

developing nations, this hierarchical division of labor may no longer be as distinct in the


IBM Global Services is the largest service provider in the world with revenues in

excess of $46 billion and 175,000 employees spread across 160 nations as of 2004.

The services it provides include application development, data storage, infrastructure

management, networking, technical support, business consulting, and outsourcing

5 According to the IBM (2005) website, its China Software Development Laboratory employed 2,000


6 There continues to be debate regarding the success of such follow-the-sun strategies. Carmel

(1999) argues that these global development projects are difficult to manage and often are

unsuccessful. On time-shifting, see Carmel and Tjia (2005).

Page 140

services. At the end of 2004, IBM employed 23,000 people in India, and an internal

planning document stated that, by the end of 2005, this would increase to 38,000;

the bulk of these employees were in Global Services. India now has more IBM

employees than any nation except the United States (Hamm 2005).

IBM Global Services is active in providing services to domestic Indian and Chinese

firms. For example, in March 2004, it signed a ten-year IT outsourcing deal for $700

million with Bharti Tele-Ventures Ltd., India's leading telecom company that included

the transfer of Bharti's IT-related assets (including workers) to IBM. Not only did

IBM acquire a new customer; it also purchased more skilled employees to expand its

Indian operations. In 2004, IBM also purchased a leading Indian business process

outsourcing firm, Daksh, which though not an IT firm, had 6,000 employees. This

acquisition illustrates how the IT and non-IT services are blurring for the providers.

For this reason, the discussion in this section incorporates an overview of IBM’s

entire range of offshoring service operations not only the software services.

India is becoming IBM’s central delivery center for services. However, like all of the

multinational service firms, IBM has also established facilities in a number of other

lower-cost nations, including China. In China, IBM Global Services has three centers,

including one opened in Dalian in 2005 with 600 workers. The Dalian center is

expected to grow rapidly with its main purpose to serve the Asia Pacific market

(ZDNet 2005). IBM Global Services also has a service center in Mexico.

In August 2005, IBM announced that it was establishing an IT services research

center in Bangalore as an extension of its India Research Laboratory located in New

Delhi with an initial staff of 10 researchers. According to P. Gopalakrishnan, the

director of IBM's India Research Lab, it would look “at how technology can improve

the capabilities and efficiency of delivery. This would include the whole spectrum of

services from infrastructure management, application maintenance, BTO to BPO

services.” (CyberMedia News 2005). If this pattern continues, India may become the

hub not only for doing offshore work but also for developing ways to automate

service delivery using software.

India has clearly become a core location for IBM to provide offshore software services,

and with the establishment of a research laboratory there to develop methodologies for the

automation of service provision, it appears as though India may become IBM’s global center

of excellence for these functions. However, all of the multinational software service

providers have a global footprint so that they can offer their customers a wide variety of

services in many different languages. IBM is likely to continue expanding its workforce in

software and other services in lower-wage nations, while growth in the developed nations is

expected to be slow.

With eight laboratories around the world (three in the United States and one each in

Switzerland, China, India, Israel, and Japan), IBM Research employs approximately 3,050

researchers. The company has steadily increased its R&D expenditures outside the United

States, from 28% in 1993 to close to 60% in 2003. In the 1990s, IBM opened three new

research labs in Austin (1995), China (1995), and India (1998). For the research

laboratories, access to the most creative individuals is the greatest priority, but it is also

true that the research centers in China and India have lower operational costs. The main

point of these research centers is to attract local talent and to conduct some of the research

on problems that are relevant to the local environment using global-class research.

There appear to be some differences in emphasis between the Chinese and Indian

laboratories. The India Research Laboratory (IRL) has about 100 researchers and focuses

on areas critical to expanding India's technology infrastructure so, while IRL researchers

work on some local issues such as text mining and speech recognition for Indian languages,

Page 141

they also work on more general research problems in the areas of bioinformatics, natural

language processing, grid computing, and autonomic computing. The IBM China Research

Laboratory (CRL) also has approximately 100 researchers. It has been working on Text-To-

Speech systems and can now provide language support for Chinese, Taiwanese Chinese,

Cantonese, Korean, Japanese, and French. It has also been working on IBM’s Websphere

Translation Server that provides machine translation between English and Chinese. In this

sense, the research profile in the Chinese laboratory is more localized.

All IBM research laboratories actively cultivate relationships with local academic

institutions. For example, the India research lab is located on the Indian Institute of

Technology (IIT), Delhi, campus where it has access to a vast pool of talent. In Israel, IBM

has built strong relationships with Haifa University and Technion. The R&D laboratories in

India and China are still quite small; however, there appears to be a commitment to

increase their size rapidly. Their missions are different: in the case of China, much of their

work will continue to be on localization and the Chinese language, while the Indian

laboratory is more likely to undertake work directly applicable to global business needs.

As the largest software/software services firm in the world in both revenue and

headcount, IBM has the most sophisticated global footprint of any firm. Not only is it

increasing its employment in developing nations in the more mundane and routine aspects

of service delivery, it is also increasing employment in software product development and

research and development. In the process, IBM’s global posture is changing from being

heavily weighted toward the developed nations to a more equal weighting globally.


Siemens Business Services (SBS) is a Siemens subsidiary that has a global practice in

performing software and other outsourced work. It employs approximately 36,000 workers

and derives substantial revenue from installing, customizing, and maintaining SAP software

in businesses. Its 2004 revenues of 4.8 billion Euros were roughly divided between

Germany (48 percent), the rest of Europe (39 percent), the United States (8 percent), and

the rest of the world (5 percent) (Siemens Business Services 2004). SBS has been under

significant cost pressure and has instituted layoffs to bring its costs under control (Blau

2005). SBS, like many other large service firms, has been globalizing its service delivery

operations and, in the process, has downsized its domestic workforce. Of SBS's 36,100

global employees, only 15,100 are now located in Germany, and 4,000 are located in its

rapidly growing Indian subsidiary.

SBS has developed a business strategy that uses a matrix of vertical industry knowledge

and sets of general competencies to serve its customers. One aspect of the matrix is the

industry expertise (vertical knowledge) or competency centers that are scattered in different

countries, for example, the paper and pulp vertical is located in Finland (Hallez 2004). The

other part of the matrix is the general activities, located in offshore sites in Canada, Ireland,

and Turkey, and they handle stabilized processes. India has two roles: it functions as a back

office operation for finance and accounting, and it does general software programming and

service and applications development for SAP programs. SBS uses Russia for very laborintensive

and repetitive back office and software application development (Hallez 2004).

Siemens also operates its Siemens Information System Laboratory (SISL) in Pune,

India.7 SISL has been involved in the development of an atmospheric disturbance

model for Boeing flight simulators, engine and auto throttle control simulation,

modeling and simulation of Weibul clutter, and GPS and INS error modeling for

measurement simulation. It has also designed a control system for wind shear

7 This section draws heavily upon Express Computer (2002).

Page 142

control on the Boeing 767, a control system for the flight management system for

the Boeing 747, and primary flight control system software as well as executing the

development of Kalman filters for GPS and INS, integrated with GPS in feed-forward

and feed-back configurations. SISL has been able to use Indian engineers to design

sophisticated software for developing-nation customers.

SBS and other parts of Siemens are interesting because, in contrast to US firms, they

place a strong emphasis on nearshoring facilities to Eastern Europe, Russia, and Turkey.

Nevertheless, SBS India is the location with the largest non-German headcount, and it

continues to grow rapidly.

Conclusions About Developed Nation Software Services Firms

Software services is in most respects a headcount and labor-cost business. The

multinational software services firms have been experiencing increasing pressure on costs

due to competition from developing-nation producers, particularly the Indian service giants

(as described later in this chapter). This has forced the multinationals to secure lower-cost

offshore labor. Both IBM and SBS are typical of other service firms such as EDS, ACS, and

Accenture in that they operate globally, but only in the last five years have they found it

necessary to build significant operations in developing nations to decrease their labor costs.

Today, the larger firms such as IBM and Accenture are rapidly increasing their headcount in

a number of developing nations, particularly India. At the same time, these firms are

holding steady on their developed-nation headcount or gradually drawing it down. Given

the ferocious competition in software services, there is little likelihood that prices will

increase substantially. This suggests that, for the large multinationals, the offshoring of

services will continue to increase in both absolute numbers and percentages of the global


Software Operations in Non-Software Firms

Today, virtually every firm in every industry sector is dependent on software. These

needs range from routine software for personal computers and small servers to more

complicated and customized software for complex and proprietary systems. All of these

systems require customization, maintenance, or updating on a regular basis. IT systems

have become an increasingly significant expenditure for businesses in developed countries,

and firms are actively trying to control these costs. One way to lower them is to offshore

the work to nations with lower labor costs.

It is difficult to even estimate the amount of software work that is offshored. Businesses

do not provide this information in their reports. If work is transferred to an overseas

subsidiary, this is an internal transfer and may remain unannounced and difficult to trace.

It is more clear who does the work. If it is not an overseas subsidiary of the company, then

it is likely to be one of two other kinds of firms. The service might be supplied by a large

service firm from a developed nation such as IBM, CapGemini, SBS, or Accenture (as

discussed in the previous section). Alternatively, the work might be outsourced to a firm

from a developing nation such as TCS or Infosys (India), Luxoft (Russia), or Softtek

(Mexico) (as discussed later). When a multinational company does the software work for its

developed-nation facilities itself in one of its developing-nation locations, it is likely that this

is not the only work done at that location. For example, as of April 2005, Dell Computers

employed approximately 10,000 people in India in a variety of tasks, one of which was to

produce software for Dell’s internal operations. The overseas operations undertake many

tasks, only one of which is software production. Having a mélange of activities can provide

the scale needed to make establishing an overseas subsidiary more attractive since the

software work may not have been of a sufficient scale to justify a subsidiary.

Page 143


In the technology sector, ATI is a good example of how a firm normally considered a

hardware firm also undertakes considerable amounts of software-related work. ATI develops

tools and technologies that sense, measure, interpret, and communicate data. The company

operates in four business areas: test and measurement, automated test, semiconductor

products, and life sciences and chemical analysis. ATI, which was separated from Hewlett

Packard in 1999, established its first Indian offshoring operation in 2001. By 2005, it had

offices in over thirty countries. Manufacturing was located in the United States, China,

Germany, Japan, Malaysia, Singapore, Australia, and the United Kingdom. ATI Laboratories

are located in California; Mizonokuchi, Japan; South Queensferry, Scotland; and Beijing,


The dot-com crash had a severe effect on ATI. At the end of 2003, revenue was $6.1

billion, down from $9 billion in 2000,9 and the number of employees had been pared

from 40,000 in 2000 to 29,000 in 2003. In addition to eliminating headcount in the

developed nations, ATI decided to establish an offshoring center in India. It already

was outsourcing some software work to India. Although the company established a

facility in India, it also decided to outsource maintenance and technical work (largely

programming) to outside vendors, while retaining strategic control.

ATI introduced what it terms the hybrid model, where outsourcing service providers

are required to operate out of its offices. This has proved to be advantageous

because it mitigates the perceived security risk of having separate leased lines from

non-ATI locations feeding into the VPN (virtual private network). It also allows ATI to

induce competition among outsourcers and minimizes transition and operational

costs, and it facilitates cross-functional communication between outsourcers and the


Work transfer has not been simple. For example, in early 2002 Agilent established a

communications software engineering group in India to automate some software test suites.

When the project encountered release delays, there was friction between the US and Indian

engineers. This was exacerbated by the dot-com crash which resulted in large US layoffs.

These difficulties slowed the transfer of additional work, and, over a period of 18 months,

the Indian team experienced a greater than 70 percent attrition rate. Despite these

difficulties, the software test suite project has expanded to include the development of new

modules and maintenance and defect correction for the entire product in India.

ATI India began with simple projects. For example, the first technical project was data

entry related to engineering services. Other initial tasks assigned to India were similarly

simple such as CAD support for engineering and quality assurance. Rather rapidly, however,

the work became more sophisticated in both the technical and administrative areas. For

example, only three years later, Indian engineers were designing application-specific

integrated circuits. The Indian engineers took on more and more R&D work in wireless

solution systems, OSS, and billing software for telecom service providers. Employment has

grown at 20 percent per year, and total employment in India reached 1,250 in March 2005.

ATI’s Indian operation is typical of those established by high-technology firms. It uses

both offshore outsourcing and developing-nation subsidiaries. ATI has established

operations and R&D laboratories in a number of nations, but India has become its largest

and most important center. Though it does the more mundane software testing and

8 The material in this case study is taken from Dossani and Manwani (2005).

9 These figures exclude the company’s healthcare business, sold to Philips in 2001.

Page 144

maintenance, the Indian operation also does more challenging work, for example,

developing software that is embedded into ATI’s core telecommunications and wireless test

equipment products. ATI is an example of a process that is underway in many hightechnology

and other industrial firms whose core products are becoming more complicated

and more software-intensive.


There is relatively little information available about offshoring of business or software

services in financial firms. What is well known is that the large money-center banks,

insurance firms, and financial firms are among the largest IT users in the world. To support

their operations, they have large internal staffs and many software service vendors. One of

the world leaders in using offshore facilities for global operations is Citicorp. It uses

outsourcing both on-shore and offshore and was one of the first firms to establish a

substantial software service subsidiary in India.10

In 1984, Citibank established its Indian software subsidiary, Citibank Overseas

Software Limited (COSL). COSL wrote software in India for Citibank’s global

operations and particularly its effort to computerize its worldwide operations

(Arthreye 2003). By the time the global computerization was completed in 1989,

COSL had developed a robust banking solution and had approximately 500

employees (Bitsaa nd). COSL used other domestic companies such as Silverline and

Nucleus Software for coding, while it handled the development of the architectural

components itself. In 1992, while COSL was being converted into a proprietary

subsidiary, two executives convinced 150 employees to follow them to form Citicorp

Information Technology Industries Limited (CITIL) which was funded by Citicorp’s

venture capital arm. CITIL did not sell to Citicorp but rather became a merchant

software firm. In 2000, CITIL was renamed I-flex. As of 2005, I-flex had 5,500

employees worldwide and over 500 customers. In August 2005, Oracle purchased a

40 percent stake in I-flex for $900 million.

The remaining part of COSL continued to work for CitiGroup. Then in 2001, COSL was

merged with another arm of Citibank, India (known as Global Support Unit (GSU)) to form

OrbiTech Solutions Ltd. which developed a suite of banking products. In 2002, OrbiTech

merged with Polaris Software Laboratories (Udani), and, by 2005, Polaris had approximately

6,000 employees, working mainly in the financial arena.

Citicorp pioneered the use of India to lower its cost of software production. From

Citibank’s initial investment in India, it spun off CITIL and COSL and apparently

today does not have large in-house software operations in India. In addition to the

software operations, Citibank also had a large service operation that did everything

from transaction processing to customer-focused call centers. In 1999, this was

spun off as e-Serve and listed on the Bombay Stock Exchange. In 2004, Citi delisted

e-Serve and brought it back in-house. As of 2005, e-Serve employed more than

10,000 workers in India. In terms of software services, Citibank was the financial

industry’s pioneer in using India and has been very important in training Indians in

software development for the global market.

Since Citicorp’s pioneering establishment of a wholly-owned software services

subsidiary in India, many other banks and financial institutions, including Deutsche

Bank (Deutsche Software), Bank of America, Barclays, ING, and JP Morgan Chase,

have established facilities in India to provide software services support for their

10 For an excellent account of Citicorp’s early Indian operations that was drawn upon heavily for this

account, see Arthreye (2003).

Page 145

global operations. Regardless of the ownership configuration, there is ample

evidence that the relative amount of software service offshoring by financial

institutions to India and possibly other locations will continue to grow. For example,

insurance firms, which thus far have been more conservative than banks, have

recently begun offshoring their IT operations.


It is difficult to be certain that offshoring will lead to a decline in the number of software

service employees in the internal IT operations of firms outside the software industry, but it

does seem possible. At ATI, there were lay-offs in the IT sector; however, the losses came

in the context of massive lay-offs because of the dot-com crash. In the current recovery

throughout the IT sector, existing firm headcount in the United States appears to be

stagnant. In other sectors, there is very little data available. For example, in financial

services, it is unknown whether the increasing headcount in developing nations such as

India has had any impact on employment in the developed nations. The most that can be

said is that non-IT firms are increasing their IT and engineering-related employment in

developing nations, and this trend is underway across many different industries, including

manufacturing firms such as Caterpillar and Nissan (Kenney and Dossani 2006).

Software-Intensive, High-technology Startups

For small startups, offshoring is often a difficult decision, although recently a number of

firms in the United States have been established with the express purpose of leveraging

lower-cost engineers offshore. For smaller firms, an offshore facility can be demanding on

management time. This is especially true because in India hiring and retaining highly

skilled individuals is difficult. In developing nations, particularly China (but also India), the

protection of intellectual property, which is usually the only asset that a technology startup

has, can be difficult. Despite these obstacles and risks, under pressure from their venture

capital backers and due to the need to conserve funds, there is ample anecdotal information

suggesting that small startups are establishing subsidiaries abroad, particularly in India, to

lower the cost and speed software development.

There is a wide variety of models for utilizing offshore skills, and the following case

studies are intended only as examples of what high technology startups are doing abroad.

These case studies are by no means exhaustive, and whether they are even representative

of current practice is uncertain. However, all of these cases indicate that engineers in lowerwage

nations can be an important resource for entrepreneurial firms.


Hellosoft is a private company established in Silicon Valley in 2000 and funded by Venrock

Associates, Sofinnova Ventures, Acer Technology Ventures, and JumpStartup Venture. It is

a growing provider of high-performance communications intellectual property for Internet

telephony (VoIP) and wireless devices. The founders are Indian-Americans who had

entrepreneurial experience in US startups, and the company was established with the

express purpose of using low-cost Indian engineering talent to create the intellectual

property that would be marketed by the US headquarters team. By plan, nearly all of

Hellosoft’s research and development is done in Hyderabad, India, where the company

employs over 100 digital signal processing engineers (Hellosoft 2005). Marketing and sales

operate out of the company’s San Jose headquarters.

The Hyderabad center develops software in areas such as 3G wireless, 802.16 (a

broadband technology), and EDGE (advanced data rates for GSM evolution). It has already

had significant research success, and, in July 2005, Hellosoft raised another $16 million

from venture capitalists which will be invested in marketing and further research and


Page 146

Hellosoft’s business plan is based on leveraging low-cost engineering talent, and the US

headquarters operates largely as an interface with the market and customers. Nearly all the

growth in technical employment will occur in India. Should Hellosoft be successful, the other

beneficiaries will be venture capital firms that may garner significant capital gains and

further relationships with other Silicon Valley service firms that assisted in the

establishment of the firm.


Netscaler was founded in 1998 to redesign a specific piece of infrastructure, the load

balancer, used in regulating Internet traffic flow. Netscaler aimed to reduce the set-up and

tear-down time for each backend server connection. After Netscaler developed a product to

demonstrate its more efficient way to handle Internet traffic, the company needed to add

other features in order to attract customers who were unsure about moving from legacy

products to new hardware that did not have industry backing. Netscaler understood that,

as long as it had the ability to see inside a connection, it could offer other on-the-fly

services. To create this ability, Netscaler hired an Indian firm known as NodeInfoTech to

help develop an on-the-fly SSL encryption engine (NodeInfoTech 2005). With the aid of

NodeInfoTech, Netscaler introduced an extension to its product, allowing the backend

servers to send unencrypted data to the Netscaler product that encrypted it and forwarded

it to the client over a secure connection.

The success with NodeInfoTech convinced Netscaler to establish an Indian

subsidiary, Netscaler India. To staff the new operation, Netscaler hired many of the

developers from NodeInfoTech (Tillman and Blasgen 2005). In 2004, Netscaler India

employed approximately 60 engineers to develop other features such as on-the-fly

compression, virtual private networks (VPNs), and integrated cache, and it planned

to double the number of Indian employees in 2005 (Hindu Businessline 2004).

Netscaler had grown to 200 total employees by 2005 when it was purchased for

$300 million by Citrix Systems who retained both the Silicon Valley and Indian


The reason Netscaler formed an Indian subsidiary was to allow the company to increase

the types of work it could do and develop tighter engineering integration (Tillman and

Blasgen 2005). Netscaler’s CEO, B.V. Jagadeesh, found that “[Indian] employees of similar

skills are as efficient as they are here. The only handicap they have against their

counterparts in the US is that they are not directly exposed to customers and customer

challenges as India is not a destination market yet. When Indian companies start to buy our

products, even that gap will be reduced pretty dramatically.” (as quoted in Tillman and

Blasgen 2005).

Netscaler continues to both offshore to its subsidiary and outsource to vendors

lower-level engineering support. With the aid of both the internal Indian and US

engineering teams, Netscaler can provide all levels of support 24 hours a day. Since

the low-level support is fully outsourced, it is hard to learn much more about its


At Netscaler, technical writing is done by in-house technical writers because it is

necessary for the writers to work closely with the engineers to provide good

documentation. Netscaler originally employed a single technical writer in the United

States, but in 2003, as the staff in India grew, a technical writer was hired there.

The company’s main reason for dividing up the writing was that the writer had to

11 This section draws heavily on a case study done by Joshua I. Tillman and Nicholas W. Blasgen


Page 147

work with the engineers in order to correctly document the various product

specifications. This allowed Netscaler to divide documentation writing between the

two development sites, and the lower wages in India allowed a net reduction in the

costs of producing documentation.

As a part of Citrix, it seems likely that Netscaler’s future growth will be divided between

the United States and India. The exact division is not yet clear, but cost pressures indicate

that Indians will become an ever greater portion of the entire workforce.


Ketera is a venture capital-financed firm established in 2000 to help firms cut purchasing

costs, streamline procurement processes, and achieve higher performance from suppliers

without the expense and overhead of traditional software applications. The company

provides its software as a service. To lower costs, Ketera made a strategic decision to use

India for all functional areas in the company. In its first phase of offshoring in 2002 and

2003, it contracted three Indian firms to provide software development, client services,

customer support, and IT support. In April 2004, the company decided to create a whollyowned

subsidiary in India and to transition from all outsource to mostly in-house offshore

operation. In 2005, Ketera has a wholly-owned subsidiary in Bangalore employing about 75

people. The company still outsources a small portion of work to a legacy provider and

contracts with new providers for special needs.

Why did Ketera set up a subsidiary? In 2004, the company was offshore outsourcing some

software development of its core service product, customer support, IT support, and some

other functions. However, the company decided that the engineers at the outsourcing firms

were not as productive and quality-oriented as Ketera desired. This problem seemed to be

due to compensation and attrition issues, and to engineers with no motivation to innovate.

There were also difficulties in the United States, where there were too few US managers to

handle the Indian engineers, resulting in significant communication gaps. These issues

prompted Ketera to establish its Indian subsidiary. Their first Indian hire was a general

manager who had experience working in both a Silicon Valley start-up and in India.

The software-related functions offshored internally were software development,

operations IT, marketing, and customer support, and portions of product management. In

2005 there was discussion of whether to move certain back office functions and

telemarketing to India. According to one report, the center was tapped to be the product

engineering and development site for the company’s entire suite of spend management

solutions (Times News Network 2004).

Shah (2005) believes that as the Indian teams mature, they will be able to perform more

sophisticated work and that other functions could be at least partly offshored. Maturation is

occurring quickly, and Ketera is already creating a new technology prototype in Bangalore.


An increasing number of US technology startups are utilizing lower-cost workers in

developing nations. These case studies indicate that, although startups may initially use

outsourcing as a strategy, they often soon opt to establish a subsidiary for a variety of

reasons, including concerns about intellectual property protection, workforce control, and

management efficiency. According to Shah (2005), the minimum staff size for an offshored

operation is about 10 people. If this is accurate, then it may be possible for many more

small firms to establish subsidiaries in developing nations. Unfortunately, data on the scale

and scope of offshoring by startups is unavailable.

12 The source for the discussion of Ketera is Shah (2005).

Page 148

It is tempting to view this offshoring as an unmitigated loss of jobs for US workers.

However, the reality is more complicated. Lowering the cost of undertaking a startup

means that the barriers to entry are lowered, and this is likely to encourage greater

entrepreneurship in the United States. The jobs created by this entrepreneurship should be

counted against those lost to offshoring. For example, Rakesh Singh, Netscaler’s General

Manager of Asia Operations, was quoted as saying, “The cost savings through outsourcing

have helped us become more competitive and experience rapid growth as a company. As a

result, we have a lot more employees in the US today than we did when we set up the India

operations” (Tillman and Blasgen 2005). So, correctly estimating the employment effect of

offshoring in the case of startups is difficult when one takes into consideration jobs created

as well as jobs lost.

Offshore IT Service Providers

The availability of capable software programmers in developing nations provided an

opportunity for entrepreneurs and existing firms to hire them and offer their services on the

global market. As discussed in Chapter 3, it was in India where this practice first began in a

significant way. Initially, in the early 1980s, because telecommunications links were not so

sophisticated, the Indian programmers were moved to the US customer’s premises. This

practice was profitable and gradually expanded and evolved as both customers and

providers became more comfortable.13 This level of comfort and the lower cost that could

be offered through remote provision of services led to a shift wherein a major portion of the

contract work was completed in the offshore offices of the contractor.

Indian firms were the pioneers in providing the offshore outsourcing of software

production and services. As Dossani (2006) shows in his case study of India, but in a lesson

that can be generalized to firms in other nations, the real explosion of outsourcing came

during the dot-com boom of the late 1990s when there was great concern about a shortage

of programmers. US firms, in particular, were concerned about the Y2K problem and sought

low-cost assistance in preparing their IT systems. These developments created an

environment where major corporations were willing to experiment with overseas vendors,

and a sufficient number of these experiments were satisfactory. The result was that

offshore vendors, particularly Indian firms, were validated as candidates for software

projects. These projects also allowed offshore vendors, again particularly Indian firms, to

grow rapidly in headcount, experience, and financial resources so that they could undertake

ever larger and more complicated projects.


Tata Consultancy Services (TCS), the largest and oldest Indian software services provider,

is an excellent example of the growth of Indian vendors (see Table1). TCS was established

in 1968 to service the in-house data processing requirements of the Tata Group and, in

1969, offered electronic data processing (EDP) services to outside clients. In 1970, it

became the exclusive Indian licensee to sell and maintain mainframe computers built by the

American firm, Burroughs. In an effort to encourage the development of an Indian computer

industry, the government enacted the Foreign Exchange Regulation Act of 1973, forbidding

13 Obviously, comfort is a subjective term that refers a person’s faith that another person(s) will

respond in certain predictable ways or that a set of agreed upon tasks will be discharged according to

a set of expected criteria. Cultural, social, economic, legal and other practices and beliefs impact our

comfort with a relationship. Comfort is increased through repeated successful interactions. As levels

of trust increase due to positive interactions, the client becomes more willing to escalate its


14 This case study draws heavily upon Dossani and Kenney (2004).

Page 149

foreign firms from operating fully-owned subsidiaries. A number of foreign firms established

joint ventures, and the Indian industry grew gradually. During this period, all of these firms

including TCS, sold and maintained computers and software systems made overseas by

their joint venture partner and offered electronic data processing services to local clients.

TCS’s overseas experience in providing software-related services began in 1974 when TCS

was asked by Burroughs to install systems at US-based clients. Burroughs was attracted by

the combination of software engineering talent and the English language skills that it had

found in the TCS workforce. This was the beginning of the body-shopping business which

entailed the dispatch of Indian programmers to the sites of overseas clients. Typically, these

assignments lasted for a few months at a time. During this period, Indian firms were

basically labor recruiters.

Table 4-1 TCS Revenues, Number of Employees and Percent of Revenues Derived

from Outside to India, 1991-2005

Year Annual

Revenues (in


Total Employees Revenue

Derived from

Abroad, %

1990 28.1 2,300 70.8

1991 45 2,600 75.6

1992 52.3 4,761 80

1993 55.9 6,450 80.7

1994 64 5,589 79.2

1995 90.1 6,071 80.5

1996 123.9 7,864 81.9

1997 169 9,929 84.2

1998 241.8 11,176 88.2

1999 357.8 12,770 89.8

2000 417.9 15,044 86.1

2001 616.2 17,607 91.3

2002 792.1 20,459 92.7

2003 1,000 24,168

2004 1,560 30,100

2005 2,240 45,700

Source: Compilation by Rafiq Dossani and Martin Kenney

As the software industry changed and Burroughs continued to lose market share, TCS

developed a growing competence in conversion work, that is, converting clients’ existing

Burroughs’ systems to work on IBM hardware. To further its growth, in 1979, TCS opened

an office in New York, the first overseas office by an Indian software firm. Entering the

Page 150

1980s, TCS remained the largest Indian software services firm. In 1980, the Indian software

industry exports were $4 million, shared by 21 firms of which TCS and a sister firm, Tata

Infotech, accounted for 63 percent. By 1984, the number of firms increased to 35 and

export revenues reached $25.3 million.

When, in 1985, TI persuaded the government to supply it with scarce satellite bandwidth,

Indian firms such as TCS also demanded telecommunications access. But it was the

acceptance of UNIX as a programming standard in the 1980s that made offshore work for

clients feasible. Again, TCS pioneered the remote project management model as it came to

be called. In 1988, only 10% of TCS’s work was done in India, but this rose to 37% in 2005.

The industry shift to UNIX and workstations also benefited Indian firms since they could

secure work converting installed applications into Unix-compatible programs. Again TCS

was a leader, but soon other Indian competitors such as HCL, Infosys, Satyam, and Wipro


The type of work TCS performed changed substantially in the 1990s. Conversion

work tapered off once most corporations completed the adoption of the common

Unix platform. This work was replaced with writing applications programs, a more

profitable activity. TCS eagerly sought higher value-added work such as systems

integration and focused considerable effort by the end of the decade into bidding for

larger projects, that is, those that required from 20 to 150 people-years. The largest

industry serviced by TCS continues to be financial services. Today, 72 percent of its

revenues continue to be in application development and maintenance, while body

shopping still provides over 60 percent of its revenue (Mahalingam 2005).

By 1991, TCS had grown to 2,300 employees and had revenues of $28 million. During

this period, the company pioneered the establishment of India-based, client-specific

offshore development centers (ODCs) which enabled firms such as TCS to undertake large,

turnkey projects that combined Indian-based and overseas staff (the latter often supplying

critical industry expertise otherwise unavailable in India). Y2K was a bonanza for TCS and

the other Indian firms. At the end of the fiscal year 2000, TCS had 15,000 employees and

revenues of $428 million. To accelerate its growth, in 2001, TCS acquired CMC, an Indian

government-owned firm with 2,500 employees. Rather than slow down after 2000, the

rapid improvement in telecommunications capabilities combined with serious pressure on

the bottom lines of firms in the developing nations expanded the opportunities for TCS

which grew to over 20,000 employees in 2003. TCS began offering new services such as

real time database management, quality assurance, and web services.

By 2005, TCS had grown to over 45,000 employees and was continuing to grow at

approximately 25 percent per year. As TCS continues its efforts to overtake firms such as

IBM and Accenture, it is establishing a global network of operations facilities, not only

marketing, customer liaison, or concentrations of dispatched personnel. In 2005, the

company had development centers in Europe, Latin America, and Japan, although most of

its employees continued to be located in India.

TCS and its major Indian competitors have had a significant cost advantage over their

developed-nation rivals. Until very recently, however, they did not have either scale or a

sufficiently global footprint to compete against the IBMs and Accentures. This is changing

as the Indian firms experience annual growth rates in excess of 25 percent and have

significantly better profitability than their US-based competitors (Hira and Hira 2005). The

marketplace dynamic may change as the rivals from developed nations increase the

percentage of their workforce located in lower-cost environments. Regardless of the

outcome, firms such as TCS have successfully forced firms from developed nations to

dramatically increase the portion of their global workforce located in developing nations,

and thereby have shifted the geography of software service provision.

Page 151


Indian firms, due to their size and sophistication, have rightfully received the bulk of the

attention from those considering offshoring. However, there are firms in other developing

nations that are also providing software services to developed nations. One noteworthy

example is Softtek, a privately-owned Mexican firm based in Monterrey with development

centers in Monterrey, Aguascalientes, and Mexico City, two others in Brazil, and one in

Spain. Like the large Indian firms, Softtek operates certified Six Sigma programs and has

reached a CMM 5 rating (Softtek 2005b). The company was established in 1982 to employ

graduates of Mexico’s best technical university, the Tecnologico de Monterrey, to provide IT

consulting services to Mexican firms and later to firms in other parts of Latin America. It

entered the US market in 1997 with the business strategy of providing a near-shore

alternative. In recent years, Softtek has grown from 2,000 employees in 2000 to

approximately 3,400 in 2005 (Lopez 2005; Softtek 2005a). With about $135 million in

revenue, it is growing at 30 percent per year, although it still is only about one-tenth the

size of providers in the large developed nations or India.

Softtek’s value proposition is based on the fact that its software development centers are

near-shore, and thus operate synchronically with its customers. Because its employees are

more highly paid than those in the Asian developing nations, Softtek had to develop a

somewhat different model than Indian vendors.15 Their business strategy is not to displace

offshore vendors, but rather to capture a portion of the total offshore spending. What

Mexico offers is an opportunity to diversify risk which is important for highly interactive

processes that could benefit from running at the same time. To further their advantage,

Softtek even adopted the US vacation calendar for their US-focused operations. In addition,

the United States and Mexico share similar cultural and commercial environments. Proximity

facilitates the logistics of arranging face-to-face meetings. This limits the need for Softtek

engineers to be stationed onsite, thus lowering costs and helping to make Softtek

competitive with the lower-cost Indian or Chinese competitors (Lopez 2005). Travel is

simplified because, when it is necessary to visit, as a Mexican firm, employees can use

NAFTA visas. In general, Softtek works on fixed-price contracts, not the time-and-materials

contracting that is typical of body-shopping.

Despite the opportunities, Mexico’s growth in the IT area has been limited. Softtek is the

largest independent Mexican software services offshoring firm serving the global market,

although there are other smaller firms. Only recently has the Mexican government

recognized the opportunity in software services offshoring and formed an organization

(Prosoft) to improve Mexico’s position by funding training projects. Even five years ago, few

Mexican universities outside of the Technologico de Monterrey were providing well-trained

graduates for this industry. This has changed as Mexican universities and students have

recognized the career potential in IT. To improve the preparation of Mexican IT workers,

Softtek and the other Mexican IT vendors are interacting with a number of Mexican

universities to improve IT training (Lopez 2005).

Softtek’s experience demonstrates that it is not only the Indian majors that are finding

opportunities to provide software services to developed nations. Yet, its status as one of

the largest Latin American software services firms indicates the lead the Indian firms have

built. This case study also shows that high-level CMM qualification is not confined to Indian

firms. Most importantly, it demonstrates the entrepreneurial opportunities available in any

developing nation that has a reservoir of technically trained personnel.

15 An IT graduate from a Mexican university starts at between $15,000-18,000 per year as opposed to

an Indian graduate that starts at $6,000 per year.

Page 152


Software services firms from a number of the developing nations are players in the global

economy. They have not yet become significant players in the packaged software industry,

and given the propensity for the large international players to buy promising software

startups wherever they may be located, it could be difficult for packaged software firms

from developing nations to capture significant global market share. The large Indian firms,

such as TCS, Infosys, Wipro, Satyam, and HCL, are at this time the global leaders.

However, in China, Mexico, and Russia, there are smaller but also rapidly growing software

service firms that employ between 1,000 and 5,000 workers. Currently, the firms from

other nations are not large enough to compete with either the developed-nation

multinationals or the large Indian firms. These medium-sized firms in other geographies can

reduce country risk for customers, although it is also possible that some of them will be

acquired. The larger multinationals and Indian firms are also establishing facilities in other

geographies, particularly Eastern Europe and, more recently, Mexico.

4.3 Overall Conclusion

The variety of case studies in this chapter illustrates the breadth of the phenomenon of

software and software services offshoring. The reasons for offshoring vary by firm and

particular recipient nation, and often decisions are made for a complicated amalgam of

reasons. In the case of the elite R&D laboratories, the desire to tap into the most talented

individuals, wherever they might be in the world, is clearly the foremost motivation.

Particularly in the case of China, but also increasingly India, the growing local markets are

attractive and a reason for siting software facilities locally. Labor costs are a primary

motivation for much of the offshoring being undertaken by the firms examined.

There can be little doubt that offshoring is still small in comparison to how large it is likely

to become. The case studies in these chapters are firms that can be considered early

adopters; the followers have only recently begun to investigate the opportunities for

offshoring. As the case of ATI showed, particularly in the subsidiaries of Western firms, it is

likely that more sophisticated work will be relocated during the coming decade. Firms are

becoming increasingly willing to entrust core activities to their offshore subsidiaries.

Whereas some believed that a certain size was necessary prior to offshoring, the case

studies of startups showed that this is not true. US startups are establishing offshore

subsidiaries even before their headcount reaches 50 people, and for some firms, their entire

business plan is built on the premise of using lower-cost offshore IT professionals. This

suggests that employment growth in the United States might be constrained. However, the

availability of low-cost technical talent also can lower the barrier to entry for

entrepreneurship, and this may encourage greater entrepreneurship and, as a result, wealth

and job creation in the United States.

Every firm in this admittedly small sample is pursuing a global strategy for R&D and IT

provisioning. It is entirely possible that this will become the norm for nearly every firm in

the developed nations. The labor-cost arbitrage factor is and will remain significant and all

executives, in large and small firms, are considering the most economical footprint for their

IT operations.

Page 153

4.4 Bibliography

Athreye, S.S. 2003. Multinational Firms and the Evolution of the Indian Software Industry.

East West Center Working Paper #51.

Barlas, D. 2004. SAP Doubles Indian Commitment. (Nov.).

default.asp?articleID=6168&TopicID=3. (Accessed Sept. 2005).

Bishop, J. 2005. China's Red Flag Software Sees Strong Growth This Year. Xinhua Financial

News (Sept. 14).

Software-Sees-Strong-Growth-This-Year.xhtml. (Accessed Sept. 2005).

BITSAA International. How They Did It.

leaders/hukku.htm. (Accessed Aug. 2005).

Blau, J. 2005. Siemens Business Services Faces Change. Network World (Feb. 4). (Accessed Sept. 2005).

Carmel, E. 1999. Global Software Teams: Collaborating Across Borders and Time Zones.

Prentice Hall, Upper Saddle River, NJ.

Carmel, E. and Tjia, P. 2005. Offshoring Information Technology. Cambridge University


Cyberhood News. 2005. IBM Opens Dedicated IT Services Research Center in Bangalore.

(Aug. 18). (Accessed Sept.).

Dossani, R. and Kenney, M. 2004. Moving Tata Consultancy Services into the Global Top 10.

Journal of Strategic Management Education 1, 2. 383-402.

Express Computer. 2002. SISL Grows Out of Parent’s Shadow. (July 1). (Accessed Aug.).

Hamm, S. 2005. IBM'S Passage to India. Business Week Online (Aug. 8).

(Accessed Aug.).

Hayward, D. 1997. How Offshore Programmers Save IBM Millions. TechWire (Feb. 18). (Accessed Aug. 2005).

Hellosoft, Inc. 2005. Corporate Home Page.

(Accessed Aug.).

Hira, R. and Hira, A. 2005. Outsourcing America. AMACOM.

IBM. 2004. IBM Increases Hiring Plans. (Aug).

ircorner/2004/12-08-04-1.phtml. (Accessed Aug. 2005).

IBM. 2005. China Software Development Laboratory.

businessunit/businessunit_7.htm. (Accessed Aug. 2005).

Indo-Asian News Service. 2005. SAP India To Hire 2,000 More Techies.,0003.htm. (Accessed Aug.).

Kumar, V. R. 2004. NetScaler To Double Headcount. (June 16).

(Accessed Aug. 2005).

Lopez, B. (Softtek CEO for US Near Shore Services). 2005. Telephonic Interview by Martin

Kenney. (Sept. 7, 2005).

Mahalingam, S. 2005. (Tata Consultancy Services). Presentation at the India Conference –

Shaping the Future (Aug.).

(Accessed Aug.).

Page 154

NodeInfoTech. 2005. Home Page. (Accessed


Orian, S. 2004. The Politics of High Tech Growth: Developmental Network States in the

Global Economy. Cambridge University Press.

Palmisano, S. 2005. 2005 Investor’s Briefing.

inv_day0505/presentation/inv_day0505prepared.pdf. (Accessed Aug.).

Parthasarathy, A. 2005. Global Sourcing—The New Reality. Paper presented at The

Globalization of Services, Stanford University (June 17).

People’s Daily Online. 2004. SAP vows to enhance R&D capacity in China. (Dec. 16). (Accessed Aug.


Kenney, M. and Dossani, R. 2005. Offshoring and the Future of U.S. Engineering. The

Bridge. (To appear).

Rediff. 2005. Rediff Interview with Naresh Gupta. (April 21).

money/2005/apr/21inter.htm. (Accessed Aug.).

SAP. 2005. SAP Bulgaria Home Page.Available at

saplabs/bulgaria/. (Accessed Aug.).

Shah, R. 2005. Ketera India Case Study. Conference ON The Globalization of Services,

Stanford University (June 17).

Siemens Business Services. 2005. Corporate Website. (Accessed


Sim, S. 2002. Adobe Says It's Committed to China Despite Piracy. (Jan. 15). (Accessed Aug. 2005).

Smith, R. 2005. Partnerships Are Important. Rediff (Aug.).

2005/aug/24inter.htm.. (Accessed Aug.).

Softtek, S.A. 2005a. Corporate Home Page. Available at (Accessed


Softtek, S.A. 2005b. Our their own words. Available at

nearshore/html/about_people.htm. (Accessed Aug.).

Tillman, J. I. and Blasgen, N.W. 2005. Case Study of Netscaler. Paper written for CRD 199

Special Study Course, University of California Davis (June 16).

Time News Network. 2004. Another IT Firm Deserts US for India. (Dec. 8). Available at (Accessed Aug. 2005).

Tsai, M. 2004. Inside DSP on Automotive Signal Processing: Feeling the Heat. Inside DSP

(Sept. 13). Available at

articleID=30900019. (Accessed Aug. 2005).

Udani, D. The Quest for HTMLParser. Available at

articles/quest.html. (Accessed Aug. 2005).

Verma, P. 2005. Adobe To Buy Macromedia's B'lore Development Centre. The Economic

Times (May 16). Available at

1111152.cms. (Accessed Aug.).

ZDNet. 2005. IBM Expands in China. ZDNet News (May 26). Available at (Accessed Aug.).

*R.C. Waters took a leadership position in the writing of Chapter 5. Page 155

Authors: Juris Hartmanis, Roli Varma, Richard C. Waters*

Chapter 5: The Globalization of IT Research

5.1 Overview

IT research has historically been, and still is, concentrated in a few countries. However,

IT research is becoming more equally spread around the globe. This globalization is almost

certainly unstoppable and may well accelerate. If current trends continue over the next

twenty to thirty years, it is likely that IT research will spread to the far corners of the world,

and China and India will emerge as centers of IT research rivaling the United States and

Western Europe.

There is little hard data on the migration of IT research jobs. However, it appears that, to

date, such migration has been limited and has on balance gravitated toward traditional

centers of IT research rather than away from them. A much more significant phenomenon

has been the migration of IT researchers themselves from one country to another. This

migration has been overwhelmingly to the traditional centers of research. The migration of

both jobs and researchers to traditional centers of IT research is lessening. The direction of

job migration may well reverse.

Globalization presents challenges to the traditional centers of IT research. If they become

complacent, or even merely inattentive, they may well dwindle in significance with strong

negative consequences for their local economies. However, the globalization of IT research

is happening in the context of a general increase in the amount of IT research. It is not a

zero sum game where increased opportunities in one place inevitably result in decreased

opportunities in other places. If they take strong action, it is entirely possible that the

traditional centers of IT research will continue to flourish even as additional centers emerge.

The Concentration of IT Research

According to data collected by the Thomson ISI science citation index for the years 1999-

2003 (see Figure 1), about a third of computer science papers come from the United States

alone. A few additional traditional centers of concentration in IT research (Australia,

Canada, France, Germany, Israel, Italy, the Netherlands, Sweden, Switzerland, and the

United Kingdom) account for another third.

Page 156

Figure 5-1: The Globalization of IT Research

The line shows where a country would be if its share of CS publications were equal to its

share of PPP GDP. The data underlying this graph are shown in Table 5-3.

Much, but not all, of the large share of the world’s IT research in these eleven countries is

explained by the large part of the world’s Gross Domestic Product (GDP) that is

concentrated in these same countries. Figure 1 plots the percentage of the world’s

computer science publications against the percentage of the world’s Purchasing Power Parity

(PPP) adjusted Gross Domestic Product (GDP) for all those countries that produce more than

1% of one or the other. There is a basic correspondence between PPP GDP and computer

science publication. However, the share of computer science publications by scientists in

the traditional centers of concentration of IT research is more than 60% greater than their

share of world PPP GDP (65% vs. 40%).

IT research was even more concentrated in the past than it is today. The initial bloom of

IT research occurred in only a few select locations in the United States and a couple other

countries in the aftermath of the Second World War. This small group of research centers

expanded shortly after to the full list of traditional research centers given previously. Over

the later 20th century, the list of IT research centers has continued to grow, but relatively

slowly. For example, in Europe, Spain, Greece, and Belgium have joined the list, and in

East Asia, Japan, South Korea, Taiwan, and Singapore have become significant research

centers. With these additions, the centers of IT research listed produce about 85% of all IT


CS publication vs PPP GDP






0.0% 0.1% 1.0% 10.0% 100.0%

% World PPP GDP

% World CS Publications












% pubs = % GDP

Page 157

China and India are moving toward becoming centers of IT research, but they are not

there yet. Some other countries with significant GDP such as Brazil, Indonesia, Mexico, and

Russia produce very little IT research. These six countries combined produce 27% of world

PPP GDP but only 7% of computer science papers.

Particularly in the United States, the initial surge of IT research was driven by ample

government funding and a significant migration of scientific talent from the rest of the

world. The continued importance of government funding is illustrated by the fact that

countries such as Israel, Singapore, and Sweden that have particularly high per capita

government funding for IT research also have particularly high levels of computer science

publication in comparison to PPP GDP. In addition, as shown in the data presented in this

chapter, there has been a general migration of scientists from countries that do not support

graduate education and research to countries that do.

Due to strong efforts to foster research on the part of a number of national and local

governments outside the traditional centers of research, IT research is slowly but steadily

becoming more global. This has been accompanied by a significant increase in the numbers

of PhDs outside the traditional centers of concentration and a reduction in the migration of

researchers to these centers. In the long run, there is no obvious reason why IT research

should be any more concentrated than world economic activity in general.

What Globalization Means for the World as a Whole

Globalization allows more and better people to participate in IT research. The growing

availability of educational opportunities around the world means that more people with

research potential are able to realize this potential, increasing the size of the IT researcher

pool and the quality of the best researchers. A freer worldwide market in research means

that potential funding for IT research can more easily be targeted to those that can most

effectively and efficiently create research results. Both of these trends increase the amount

of scientific advancement that can be obtained from a given level of resources. There is

little doubt that this is good for the field of IT and for the world as a whole; however, while

we gain as a group, there can be individual losers.

What Globalization Means for Individual Locations

Research, in general, and IT research, in particular, is one important foundation for high

value-added economic activity and is actively sought by more and more locations. This

chapter uses the word location instead of country to highlight the fact that issues of change

in IT research activity are not tied to countries so much as to particular regions within

countries. For example, inside the countries that are the traditional leaders in IT research,

there has long been competition between established research locations and new locations

wishing to achieve that status. This competition is little different and no less intense than

the global competition that is now emerging.

Becoming (or maintaining one’s status as) a center of research in any field requires

consistent long-term effort. The required measures include building basic economic

infrastructure, providing first-rate education through the doctorate degree level to train high

quality researchers and attract first-rate students who stay in the location, and providing

ample direct government funding for research as demonstrated by the data presented in

this chapter.

Every location must realize that it is competing in a truly global marketplace. This

presents opportunities for locations that are not yet centers of research and challenges to

those that are. It is likely that the traditional centers of concentration of IT research will

remain important centers of research because as significant research centers, these

locations will naturally attract research funding and research talent. However, these centers

must take continued active measures to foster research. They cannot be complacent and

Page 158

assume that merely being a center of concentration of IT research is, by itself, a guarantee

of indefinite success.

What Globalization Means for Individual Researchers

Globalization provides improved opportunities for people who live outside the traditional

centers of concentration of IT research. It also provides improved opportunities for the best

researchers due to increased global competition for their services. However, it limits the

opportunities of the least skilled researchers in the traditional centers of concentration, for

whom global competition may mean declining wages or even the loss of jobs.

Every researcher must realize that he or she is competing in a truly global marketplace.

There are many people worldwide who could be good IT researchers. Among those who are

already researchers, huge differences in skill exist, and this translates into large productivity

differentials. Those with talent who pay attention to maintaining a high skill level should

see opportunities from globalization, but they must realize that they can no longer fall back

on merely living in a traditional center of concentration of IT research as a guarantee of

indefinite success. Because of the higher quality and productivity that results, talent and

skill level will eventually win out wherever it is to be found globally.

Worldwide Changes in the Balance of Supply and Demand for IT Research

The globalization of IT research will inevitably reduce the dominance of the traditional

centers of concentration in relative terms. However, IT research is not a zero sum game.

The most important question for individual locations and researchers is not whether they

will prosper in comparison to others, but whether they will prosper in comparison to their

own past history. If a given location has a vibrant and growing IT research community, it

matters little if other locations are growing more rapidly. Similarly, if a given researcher

has a career that is growing in interest and pay, it does not matter much if the prospects of

other researchers are increasing more rapidly.

This chapter is primarily about changes in the balances between locations. If the demand

for IT research and the supply of IT researchers were static, then this would be a primary

determiner of the future prospects of locations and the researchers in them. However, the

situation is far from static.

Both the demand for IT research and the supply of IT researchers are increasing rapidly.

The most important question of all is whether the demand or the supply is increasing more

rapidly. Changes in the worldwide balance of supply and demand for IT research is a more

important factor for predicting the future than changes in the balance between locations.

Unfortunately, forecasting the future balance of supply and demand comes down to

forecasting the difference between two large, rapidly growing, and hard to forecast

numbers—a very difficult task.

The goal of IT research is the automation of information and knowledge manipulation

tasks, and as such, it is arguably one of the most fundamental of all disciplines, contributing

to every area of science, engineering, and the economy. There is therefore every reason to

believe that the overall demand for IT research will be very strong—quite possibly strong

enough to grow faster than the worldwide supply of quality researchers.

Why IT Research Is a Separate Section in This Report

Discussion of research is in a separate section of this report because it is a self-contained

microcosm with product flows that are quite different from IT in general. In addition, the

indicators of what is happening in worldwide research, such as the publication of research

papers and the numbers of PhDs, are different from the indicators of IT development

activities. However, developments in the globalization of research may well be fundamental

harbingers of changes to the field as a whole.

Page 159

The Lack of Direct Data

It would be advantageous to start with a clear definition of what IT research is and then

collect a set of data that directly targets that definition. However, there is little available

data that directly targets any definition of IT research. Rather, data typically lumps IT

research with other kinds of research, advanced development, or both. For instance, much

of the data from the National Science Board combines all of natural science and engineering

together. Similarly, economic data on the IT industry typically lumps research

expenditures with advanced development costs and often with other things as well.

As a result, we see little advantage in arguing for any particular definition of IT research.

Instead, we present a range of data relating to IT research. No single piece of this data is

authoritative in its presentation of what is happening in IT research. However, since every

piece of data paints a qualitatively similar picture of steady globalization, we are confident

that this picture substantially applies to any plausible definition of IT research.

5.2 Worldwide Distribution of IT Research

Insight into the distribution of IT research can be gained by looking at R&D expenditures,

the publication rates of research papers and patents, the international ranking of

universities, and the granting of doctoral degrees.

Overall R&D Expenditures

As shown in Figure 2 from the National Science Board’s (NSB) Science and Engineering

Indicators for 2004, worldwide research and development is concentrated in a few

industrialized nations. Of the $603 billion in estimated R&D expenditures in the year 2000

for the thirty OECD countries, fully 85% is spent in only seven countries (Canada, France,

Germany, Italy, Japan, United Kingdom, United States) and more than 40% in the United

States alone.

Note that all the curves in Figure 2 are trending upward but that research is growing

fastest in the countries that currently do the least research. Continuation of these trends will

inevitably lead to a more equal distribution of research around the world.

Figure 3 from the NSB (2004) shows R&D expenditures as a percentage of GDP.

Comparing the two figures reveals that most of the differences in R&D spending stem from

differences in GDP. However, within the G-8 countries, non-defense research and

development as a percent of GDP differs by a factor of three between the lowest and

highest. It is interesting that these differences have been quite stable over the past twenty

years. In comparison to Figure 3, China spends only 1% of its GDP on research, and some

small high-tech powerhouses, including Israel and Sweden, spend in the range of 4% and

more (see the NSB (2004, Table 4-17)).

These figures aggregate data on many kinds of research and development. Consider the

following more detailed information about US government funding of research and

development. The US National Science Foundation (NSF) (see James (2005) reports that

US government R&D funding dropped from 1.25% of GDP in 1985 to only 0.75% of GDP in

2002. Over this time, research and development in the life sciences remained more or less

flat at 0.41% of GDP, but funding for research on technology dropped precipitously, from

0.55% of GDP to 0.24%. As a result, while overall research and development is rising in

the United States, the government is not emphasizing technology research nearly as much

as in the past. This change of emphasis in the United States is likely to accelerate the

globalization of IT research.

Page 160

Overall Research Publication

The US National Science Foundation compiles statistics on the publication and citation of

Science and Engineering (S&E) papers in general. Figure 4 from the NSB (2004) presents

the output of S&E articles for various regions and countries.

The picture painted by Figure 4 is broadly similar to the one painted by Figure 2 but

focused more on IT research. The principal S&E research contributors are Western Europe

and the United States. There has been steady growth of Western Europe’s research output

that overtook the US output in numbers of publications in the mid-1990s. The research

output from Emerging East Asia (Taiwan, South Korea, Singapore, and China) is small but

growing rapidly. Given the economic vitality and the strong growth of S&E PhD degrees in

this region, it is reasonable to expect East Asia to emerge as a strong new research region.

In addition to the publication of papers, NSF tracks the citations to these papers. This

correlates with the quality and influence of the papers coming from various regions which is

much more important than mere numbers. This data is summarized in Figure 5 from the

NSB (2004).

Here the dominance of the United States is greater, but the same picture of slow and

steady globalization emerges. Western Europe is steadily catching up with the United

States. Papers from Japan are cited approximately half as many times on average as US

papers, and to date papers from Emerging East Asia have even less influence. It will

probably take a long time for Asia to catch up with the United States, but it is in the process

of doing so.

Another way of assessing the influence of S&E research from various countries is by

considering the number of citations in US S&E literature, shown in Table 1 from the NSB

(2004). The absolute level of citations may not be all that meaningful because US

researchers are more likely to read and cite articles written in English and because they are

perhaps more likely to read articles from researchers located geographically close to the

United States. However, the relative level compared to other countries should have

meaning. The strong stability of the citation percentages of the countries shown suggests

that the importance of the research in these countries has changed little on a relative basis

between 1994 and 2001.

University Rankings

A large portion of research is carried out in universities, and much of the best research is

performed at the best universities. Insight into the distribution of the highest quality

research can be obtained from the distribution of the world’s best universities. As

demonstrated in Table 2, the distribution of the top 100 universities in the world has the

same basic form as the distributions in Figure 5 and Table 1. (The data in Table 2 are

based on a list of the world’s best universities compiled by Shanghai Jiao Tong University in


IT Research Publication

Table 3 shows the percentage of the world’s computer science publications (as compiled

by the Thomson ISI science citation index for the years 1999-2003) along with the

percentage of the world’s Purchasing Power Parity (PPP) adjusted Gross Domestic Product

(GDP), for all those countries that produce more than 1% of either. The table uses PPP GDP

rather than nominal GDP because the primary expenses of computer science research are

salaries, and PPP GDP is more closely aligned with salary costs in a country.

Unsurprisingly, there is a strong correlation between computer science publications and

PPP GDP. However, there are important deviations from this correlation. The principal

centers of IT research (United States, Japan, United Kingdom, Germany, Italy, France, and

Page 161

Canada) generally produce considerably more computer science publications than would be

expected from their PPP GDP alone. Some smaller countries including Taiwan, the

Netherlands, Greece, Sweden, and Switzerland produce more than twice as many

publications as would be expected from their PPP GDP. Singapore and Israel produce 7 and

8 times as much, respectively.

At the other end of the spectrum some countries with substantial PPP GDP (e.g., China,

India, Russia, and Brazil) produce relatively few computer science publications. Mexico

produces less than one ninth of what would be predicted by PPP GDP, and Indonesia

produces almost no computer science publications at all.

From Table 3, it is clear that the correlation of computer science publication is not just

with PPP GDP, it is also with leading-edge, high-value-added economies.

PhD Degrees Conferred

The number of S&E PhDs conferred is an indicator of a region’s research effort because

much of the world’s IT research is done at universities by doctoral students. In addition, the

number of computer science PhDs is a key factor supporting a region’s future ability to

perform research because highly trained researchers are the most important foundation for

research. Figures 6 and 7 from the NSB (2004) show the rate of Natural Science and

Engineering (NS&E) PhD degrees awarded for selected countries.

Particularly striking in these graphs is the recent huge growth of NS&E PhDs in Asia, in

general, and China, in particular. This contrasts with the United States and Germany,

where strong growth in the 1980s has given way to decline, and also other countries, where

there has been steady growth for many years. Changes in the number of PhD degrees

suggest that research output will soon rise in East Asia, while stagnating at best in the

United States and Germany.

These data are for NS&E PhDs as a whole. Looking more specifically at computer science

PhDs, the data is not as comprehensive but suggests similar trends. According to the NSB

(2004), there were 7,389 PhDs awarded in mathematics and computer science lumped

together in 2000. Of these, 1,832 (24%) were in the United States, while 4,057 (55%)

were in the European Union, with 956 in Germany, 800 in France, 760 in the United

Kingdom, and 704 in Italy. This data is difficult to interpret because Europe has a higher

proportion of mathematics doctorates than the United States, and the data set is missing

information about countries in Asia. As a result, the US share of computer science PhDs

may well be higher than the US share of NS&E PhDs as a whole.

Figure 8 is taken from the Computing Research Association (CRA) (2004) and shows that,

while there has been a bit of an up-tick in the past year, the number of computer science

PhDs in the United States has been basically trending downward for many years. Other

data from the same survey shows increases in the number of students passing PhD

qualifying exams, which suggests that the recent higher level of PhDs may continue.

Nevertheless, Figure 8 still stands in marked contrast to the vast increase in graduate

education in places such as China and India.

The Big Picture in Research Distribution

The previous data all indicate that the United States has the world’s preeminent S&E

research effort, followed at some distance by the United Kingdom, Germany, France, and

Japan. Looking more specifically at IT research, some smaller countries such as Israel,

Singapore, Taiwan, Greece, Sweden, Switzerland, Canada, and the Netherlands stand out as

producing a large amount of research in comparison to their size.

The data showing trends over time all indicate that the preeminence of the United States

and Europe is waning, and the gaps between countries are narrowing. It is not a question

Page 162

of whether these gaps will narrow significantly, but when. In particular, the data on PhD

degrees conferred indicates a rapid narrowing.

For example, if the trends in Figure 6 continue, the number of PhD degrees in China will

equal current US levels in 15 years or so. The output (and particularly the impact) of

science from China is not yet rising as quickly, but this is not surprising given the

assumption that the number of PhD degrees awarded is a leading indicator of scientific

output, and the increase in output has not yet fully responded to the major acceleration in

Chinese PhD degrees that started ten years ago.

Unless something seriously derails current trends, it seems almost certain that China will

be a research center rivaling the United States and Western Europe in importance within

twenty to thirty years. The development of critical scientific infrastructure in India is a few

years behind developments in China but moving down a similar road.

Figure 5-2

Page 163

Figure 5-3

Page 164

Figure 5-4

Page 165

Table 5-1: Countries whose S&E articles were cited most in U.S. S&E articles: 1994

and 2001

Table 5.2: University ranking. Data taken from Institute for Higher Education,

Shanghai Jiao Tong University, 2004, Academic Ranking of World Universities.

Number of top-100 universities in countries

In 1st 25 In 2nd 25 In 3rd 25 In 4th 25 In top 100

US 18 17 10 6 51

UK 4 1 2 4 11

Germany 1 2 4 7

Japan 2 1 2 1 6

France 2 2 4

Sweden 1 1 2 4

Canada 1 1 1 3

1994 2001

Rank Country Percent Country Percent

1 United Kingdom 17.8 United Kingdom 16.0

2 Japan 12.4 Germany 12.7

3 Germany 11.9 Japan 11.9

4 Canada 10.4 Canada 8.9

5 France 9.2 France 8.7

6 Netherlands 4.5 Italy 5.1

7 Italy 4.2 Netherlands 4.5

8 Switzerland 3.9 Australia 3.9

9 Sweden 3.7 Switzerland 3.8

10 Australia 3.7 Sweden 3.2

NOTE: Countries ranked by share of foreign S&E literature cited in U.S.-authored scientific articles.

SOURCES: Institute for Scientific Information, Science Citation Index and Social Sciences Citation Index; CHI

Research, Inc.; and National Science Foundation, Division of Science Resources Statistics, special tabulations.

Page 166

Switzerland 1 1 1 3

Netherlands 1 1 2

Australia 1 1 2

Denmark 1 1

Finland 1 1

Norway 1 1

Russia 1 1

Austria 1 1

Israel 1 1

Italy 1 1


Table 5.3:. The %CS publications numbers are from the Thomson ISI science

citation index for the years 99-03. The Purchasing Power Parity GDP numbers are

from the US CIA world handbook 2004, see (This

data is graphed in Figure 5.1.)

Country % CS pubs % PPP GDP

United States 32.3% 21.3%

Japan 7.3% 6.9%

United Kingdom 5.9% 3.2%

Germany 5.9% 4.4%

Italy 4.4% 3.0%

France 4.4% 3.2%

Canada 4.2% 1.9%

China 3.9% 12.5%

South Korea 3.4% 1.7%

Taiwan 3.3% 1.0%

Australia 2.1% 1.1%

Netherlands 1.9% 0.9%

Spain 1.9% 1.7%

Israel 1.8% 0.2%

Singapore 1.4% 0.2%

Greece 1.2% 0.4%

Page 167

India 1.2% 5.9%

Sweden 1.1% 0.5%

Switzerland 1.1% 0.5%

Belgium 1.0% 0.6%

Russia 0.8% 2.5%

Brazil 0.7% 2.7%

Mexico 0.2% 1.8%

Indonesia 0.0% 1.5%

Figure 5-5

Page 168

Figure 5-6

Page 169

Figure 5-7

Figure 5.8: PhD Production. 19 forward to fill the gap.bringFrom the 03-04 CRA

Taulbee survey.

Page 170

5.3 Researcher Migration

The knowledge of scientists and engineers can be transferred across national

borders easily through the physical movement of the scientists and engineers

themselves. This movement can be for a short term or involve permanent migration.

Since the beginning of IT research, the permanent in- migration of scientists and

engineers from other countries has been a vitally important basis for the dominance

of the traditional centers of research. The primary reason for this migration has been

people moving in search of better job opportunities. This is aided and abetted by

multinational corporations that vigorously recruit permanent employees from


Consider the movement of researchers to the United States as an example. The

National Science Board (2004) reported that, in April 1999, at least 27% of S&E

doctorate holders in the United States were foreign born, along with 20% of those

with S&E master’s degrees and 10% of S&E bachelor’s degree holders. These

individuals came from a wide range of countries around the world; however, India

and China provided the greatest number of transplanted researchers, particularly for

PhD-holding immigrants, 20% of whom come from China and 16% from India.

Mobility of Inventors

Manuel Trajtenberg (2004) of Tel Aviv University has done a study of the

movement of inventors of US patents between countries during the period 1975 to

1999. He looked at the 650,000 people who are inventors on more than one patent

and calculated statistics based on the country they were in when each patent was

filed. There were only 20,767 inter-country moves recorded in the data. This means

that 3% or less of these inventors are known to have moved. Nevertheless,

interesting patterns are evident in their movements.

Figure 9 shows the number of moves per year for each 10,000 US patents filed.

The mobility of inventors has increased steadily and markedly, rising eightfold in

twenty five years.

Figure 10 tabulates the countries these inventors moved between. The United

States had by far the largest net immigration of these inventors. Trajtenberg (2004)

argues that the total turnover of inventors to and from a country is even more

significant than the net migration because the ebb and flow of people and their ideas

is a vital stimulus to research

Page 171

Whether or not Trajtenberg’s conjecture is true, the increasing mobility of inventors is a

clear indicator of the increasing globalization of the market for scientific talent.


A major factor in technical migration is students who relocate to study in universities and

then remain in the countries where they obtain their degrees. The United States is the

most common destination for such students, but a number of other highly developed

countries (e.g., in Western Europe) are the targets of significant numbers of students as

well. The great importance of this migration of technical talent on research in the United

States is discussed in a recent report by the US National Academies (2005).

In 1997, 66% of the people in US universities who received PhDs in computer science

held student visas (see the NSB (2004, Table 3-28). By 2001, this number had decreased

slightly but was still 63%. These numbers are particularly important because many of these

students stay permanently in the United States.

According to Michael Finn (2003) of the Oak Ridge Institute for Science and Education),

56% of 1996 US S&E doctoral degree recipients with temporary visas remained in the

United States in 2001. The number of foreign students staying after obtaining their

doctorates implies that approximately 3,500 foreign students remain from each annual

cohort of new S&E doctorates in all fields. Stay rates differ by field of degree, ranging from

26% in economics to 70% in computer and electrical engineering.

As shown in Figure 11 from the NSB (2004), there has been a significant decline in foreign

students coming to the United States in recent years. One can speculate that this is partly

due to the restrictive visa atmosphere following the events of September 11, 2001. (The

refusal rate for F-1 student visas has risen from 28% to 35%, and the application rate has

fallen by 18%.) However, other forces are at work as well. This issue is discussed further

in Chapter 8.

International competition for high-quality graduate students is increasing as both

advanced and advancing countries seek more foreign talent. Job opportunities are also

becoming more widespread in the world. As a result, students now have more choices of

where to go to study, and they have more opportunity to stay in or near their home


Page 172

Figure 5.9: Moves from one country to another for inventors of multiple patents,

normalized by the number of patents filed (moves per 10,000 filings).

Figure 5.10: Total flows of inventors between countries.




















Moves per 10,000 fillings

Country Moves in Moves out Net Turnover

Canada 1392 1554 -162 2,946

Switzerland 702 693 9 1,395

Germany 1551 1701 -150 3,252

France 665 665 0 1,330

UK 2181 2809 -628 4,990

Israel 248 219 29 467

Italy 205 186 19 391

Japan 1114 1244 -130 2,358

Korea 371 270 101 641

Netherlands 453 527 -74 980

Taiwan 275 176 99 451

US 8041 7272 769 15,313

Page 173

Figure 5.11

5.4 Research Job Migration

The question of exactly what is IT research job migration is fraught with complexity. The

standard definition of job migration is that a job migrates from country X to country Y when

a company C fires a worker in X that was making product used in X and then hires a worker

in Y to produce the same product for use in X. In particular, it is not considered job

migration if C hires workers in Y to produce product to be used in Y. It is difficult to apply

this definition to research and there are questions surrounding this standard definition that

are particularly pointed from the perspective of research.

For one thing, unlike manufactured goods, there is little if any information about where

companies that create research use it. One could say that this issue is not as relevant to

Page 174

research as to other kinds of economic activity, but it does not seem reasonable to say that

it is totally irrelevant. If a company C opens a lab in China in order to experiment with

human-computer interfaces supporting the Chinese language so that C can sell more

product in China, is that job migration?

In addition, for much of the history of IT research, the research workforce has been

growing in every place where IT research is done. Is it job migration if the workforce in one

geographical location merely grows more slowly than it might have? Is it job migration if

the only alternative to moving the job from country X to country Y would have been moving

a person from country Y to country X to do the job? Here too, there does not appear to be

anything other than anecdotal information about what is actually happening.

Due to these difficulties, this section focuses primarily on where research is done rather

than on whether jobs have migrated; however, Figure 12 from the NSB (2004) shows that

the balance of trade in research is such that the United States exports more research than it

imports. Figure 13 from the same report shows that most of the research investment flow

into the United States comes from other traditional centers of research concentration and

most of the investment outflow goes to these other centers.

These data do not directly address the question of job migration, but they suggest that, to

the extent there has been job migration, it has probably been to the United States rather

than from it. Since the United States is one of the most expensive places in the world to do

research, this job migration is clearly not motivated by a search for low-cost labor.

Why Companies Do Research in Remote Locations

Before considering why companies do research in remote locations, it is important to note

what kind of research companies do in their remote locations. There are numerous

examples of companies that have moved their primary manufacturing, or even all their

manufacturing, to distant places. However, there are very few examples of companies that

have done that with research. In general, distant research labs are relatively small satellite

operations focusing on specialized areas. That is to say, companies that have distant labs

typically have much larger labs in their home areas that are the backbone of their research.

Focusing on IT research in particular, there are anecdotal reports of recent start-up

companies in California that have all of their technical operations, including research, in

India. However, other than that, we are not aware of any company in the IT business that

has a primary research lab (as opposed to a satellite lab) in a distant location nor are we

aware of any company in the IT business that is thinking of opening such a lab. It seems

entirely likely that there will be primary IT research facilities in places such as India and

China, but that will be because these places will have major IT companies that chose to

have primary IT research facilities in their home areas just as Japanese companies in the IT

business chose to do decades ago.

It is useful to distinguish two quite different cases of companies opening labs or utilizing

independent research labs in distant locations: (a) companies opening research labs in the

traditional centers of IT research concentration and (b) companies opening labs in other

locations. As noted previously, it appears that to date (a) has been more common than (b).

However, it is hard to imagine that (b) will not also be important.

Dalton and Serapio (1993, 30) present an interview survey of senior R&D executives of

Japanese electronics companies, which found the following to be important reasons to open

research labs in the United States (in no particular order).

1. Keep abreast of technological developments.

2. Help the parent company decide what technology to acquire.

3. Cooperate with other US R&D labs.

4. Hire US scientists and engineers.

Page 175

5. Assist the parent company in meeting US customer needs.

We think this list is entirely reasonable. Note that items 1-3 are central reasons for

placing research in an existing center of concentration rather than somewhere else, and

they are likely to be key reasons why such centers are self-perpetuating. Item 4 could be

interpreted various ways, but, in this case, it can be assumed to focus on exploiting the

talent pool in the center of concentration. (Given that US wages were higher than Japanese

wages in 1999, It certainly was not an attempt to save on labor costs.) Item 5 is the only

item unrelated to the fact that the United States is a center of IT research concentration.

Turning to case (b), here is a comparable list of reasons for investing in research outside

the traditional centers of concentration.

1. Take advantage of local offers of cost sharing.

2. Meet local demands for research investment.

3. Hire local scientists and engineers.

4. Assist the parent company in meeting local customer needs.

As discussed in the following, items 1-2 are the result of locations working to attract

research to their shores. Item 3 typically involves hiring lower cost labor. As a result,

items 1-3 are all indicative of job migration. In contrast, item 4 is identical to the final

reason in the list of reasons for case (a). To the extent that it is a dominant reason for the

investment, the investment is not job migration in the standard sense.

It appears that research job migration from the traditional centers of concentration to

places such as India and China is beginning to become a significant factor. In particular,

quite a few R&D labs have been created recently in these countries. However, it is very

difficult to pin down how much research job migration has actually occurred because it is

very hard to determine how much of the work done in these new labs is actually research as

opposed to advanced development.

An interesting model has emerged for staffing labs in places such as China and India

where many of the employees are hired locally at wages determined by the economy of the

host country, but the key lead research positions are filled with people brought in from

outside. Typically these lead researchers are people who grew up in the host country but

who were educated in the traditional centers of IT research concentration and gained key

research experience there. (For example, the founding head of Microsoft’s research lab in

Beijing grew up in China, got his PhD at Carnegie Mellon University in 1988, and worked for

ten years in the United States before being hired by Microsoft to start their new lab in

1998.) In addition to the natural cultural and familial attractions of returning to their

countries of birth, these people are induced to return in part by offering them salaries that

may be low by US standards but extremely high by the standards of the local economy.

Making research pay off for a company is difficult, and there is no doubt that this is made

even more difficult when a lab is located far from the main operations of the company.

However, for the most part, it appears that companies are satisfied with their overseas

research operations. Perhaps the strongest indicator of this satisfaction is the longevity of

many overseas labs. This is particularly true for case (a) discussed previously, where many

labs have a long track record. Given that it typically takes a number of years before any

newly-created lab has a real impact on the company that creates it, much remains to be

seen about the labs being created now.

Why Locations Seek to Foster Research Activities

For a country to have companies that are at the forefront of innovation is generally seen

as essential for robust economic growth in the long-term. To be at the forefront of

innovation, a location must have access to cutting-edge research and have a workforce

Page 176

capable of utilizing it. Fostering research helps both of these prerequisites. It creates

cutting-edge technology and it hones the skills of cutting-edge personnel.

The importance of research in and of itself is demonstrated by Figure 14 which shows nine

industries, each worth at least a billion dollars, spawned by IT research. Research

contributed to each of these fields in the early stages of their development. In these

important cases, government funding was critical to funding the research and establishing

the industry. (In some other important cases, industry provided the initial funding.) In the

cases described in Figure 14, the initial research phase was followed by industrial research

and culminated in a new industry in the sponsoring country. It is beyond the scope of this

study to untangle the complex interplay between basic research, customer requirements,

product development corporate research, and government. The main point is that research

is a driver of major economic development, and government funding has historically played

an important role in priming these developments.

Creating cutting-edge personnel is probably just as important as creating new technology.

Even if a location would be happy just to import research to incorporate into products it

makes rather than to import whole products from other areas, importing research is easier

to talk about than to do.. To import and effectively use research, you have to have people

that understand it fully. One of the best ways to do this is to have a research lab that is

participating in the research area because researchers in the area are in an optimal position

to find out about and understand what is happening at the research frontier of that area.

Typically, the goal of a location is not research job immigration but rather the positive

benefits to be obtained from homegrown research used at home. The end goal is a vibrant

local industry fueled by local research rather than being an exporter of research. When a

location fosters research, it has an important goal focused on job creation. However, this

goal is focused on the many jobs that can be created by a general increase in economic

activity that is sparked by research rather than on the relatively few jobs that are involved

in the research itself.

What Is Needed to Foster Research in a Country

Quality researchers and the money to hire them are critically important in fostering

research. High-quality equipment along with a high-quality communication infrastructure is

also required, but, in contrast to many other areas of science and engineering, IT

equipment and infrastructure have relatively low cost. Moreover, battles in the marketplace

during the dot-com boom led to a world-spanning broadband communication infrastructure

that is widely (though not universally) available with costs driven rapidly down because of

excess capacity. Without any connection to products or product development, it is hard to

visualize good research except in the most academic sense. For example, much of the Xerox

PARC work and the IBM work on relational database and RISC technologies, both seminal

efforts, were driven by a desire to introduce new products. While equipment,

communication infrastructure, and relation to product development are all important to

research, we will focus here on the importance of personnel.

To host research, a location needs to produce, retain, and attract quality researchers. To

produce quality researchers, a location must have first-class education through IT graduate

school. To retain quality researchers, a location must have a good work and living

environment, and good opportunities for researchers. To attract quality researchers to move

to a location, the location must have a very good work and living environment, and very

good opportunities for researchers.

The traditional centers of concentration of IT research have prospered in a self-reinforcing

way by being among the world’s best places for education, work and living environment,

and researcher opportunity. Multiple reinforcing cycles perpetuate this. The presence of

good research in universities both improves graduate education and attracts better

Page 177

students. These two factors act to produce better researchers. Researchers have a

tendency to stay where they are educated. The more research there is in an area, the more

opportunity there is for researchers. Research leads to increased economic activity, which

improves work and living environments.

Locations that want to become centers of research concentration need to invest in

improved education and infrastructure as well as direct support for research. In some cases

they also need to induce foreign companies to open research centers, for example by

offering tax incentives or by making it a requirement of doing business there.

As discussed in Newman et al. (2004), many countries are investing large sums in higher

education. In addition to this expansion of homegrown universities, some US universities

(particularly for-profit ones) are beginning operations in other countries. All told, in the

world as a whole, the number of students studying in college and graduate school more

than doubled from 40 million in 1975 to 80 million in 1995 and is continuing to grow rapidly.

An interesting aspect of IT research is that the largest traditional centers of concentration

are all in English-speaking countries, so English is very much the common language of IT

research. As a result, it is of benefit for a location seeking expanded IT research to speak

English (at least for work in IT). For instance, some German universities are now teaching

all their IT classes in English in order to provide better opportunities for their students.

There are long lead times in the various steps mentioned in this chapter, so the rate of

change is slow. Patient application of resources is required over decades before the

reinforcing cycles discussed can come into play. However, there is ample evidence that a

location can make strong progress given sufficient time and effort. This can be seen, for

example, in the experiences of many state university regions in the United States such as

the Research Triangle in North Carolina.

Particularly notable are small countries (including Switzerland, Sweden, Israel, and

Singapore) that have historically supported research to a high degree and reaped ample

rewards from doing so. For instance, Sweden has consistently provided some of the world’s

highest per capita levels of government support for higher education (currently 0.8% of

GDP, more than twice US levels) and research (1% of GDP, nearly twice US levels). This

has yielded consistently high levels of research as demonstrated by per capita publication

rates that are among the highest in the world (nearly twice US levels) and other criteria

(see Vinnova (2004)).

There is ample anecdotal evidence showing the benefits that accrue to a location that

fosters research. Given the large amounts of money and effort being expended by many

countries, there is little doubt that they feel that this is very important. This may well be a

prime area of competition between countries in the 21st century.

Page 178

Figure 5-12

Page 179

Figure 5-13

Page 180

Figure 5-14

Page 181

5.5 Conclusion

IT research is steadily, and almost certainly inevitably, becoming more global. This will

bring strong advantages to those locations that are now entering the IT research mainstream.

Because this is happening in the context of a general worldwide growth in IT

research, these benefits will not necessarily come at the expense of the current centers of

IT research. However, these current centers are faced with an important choice. They can

continue to be strong supporters of IT research and compete vigorously in which case they

should be able to continue as influential centers of IT research. However, if they choose to

ignore the growing global competition, the world may pass them by and relegate them to

second-class status.

5.6 Bibliography

James, J. 2005. Losing the Competitive Advantage? The Challenge for Science and

Technology in the United States. American Electronics Association.

National Science Board. 2004. Science and engineering indicators 2004. National Science


Computer Research Association. 2004. 2003-04 Taulbee Survey.

Institute for Higher Education, Shanghai Jiao Tong University. 2004. Academic Ranking of

World Universities.

Committee on Policy Implications of International Graduate Students and Postdoctoral

Scholars in the United States, Board on Higher Education and Workforce. 2005. National

Research Council. Policy implications of international graduate students and postdoctoral

scholars in the United States.

Trajtenberg, M. 2005. Retaining the competitive edge in the era of globalization. The IEEE

International Conference on Software - Science, Technology and Engineering. Herzlia Israel


Finn, M.G. 2003. Stay rates of foreign doctorate recipients from US universities 2001. Oak

Ridge Institute for Science and Education.

Dalton, D. and Serapio, M.. 1993, Foreign R&D facilities in the United States. Research

Technology Management.

Newman, F., Couturier, L., and Scurry, J.. 2004. The New Competition, In The Future of

Higher Education: Rhetoric, Reality, and the Risks of the Market. John Wiley & Sons.

VINNOVA (Swedish Agency for Innovation Systems). 2004. The Swedish National

Innovation System 1970-2003: A Quantitative International Benchmarking Analysis.

Page 182 *S. Goodman, C.H. House, & R. Ramer took a

leadership position in the writing of Chapter 6.

Authors: L. Jean Camp, Seymour Goodman*, Charles H. House*, William B. Jack, Rob

Ramer*, Marie Stella

Chapter 6: Offshoring: Risks And Exposures

6.1. Introduction

In June 2005, the news media reported that some 40,000,000 credit card accounts at

CardSystems of Phoenix, AZ, had been compromised by an infiltration. “The intruder

gained access to names, account numbers, and verification codes critical for committing

fraud. A MasterCard spokeswoman said the company was aware of information being

removed from the CardSystems database on about 68,000 MasterCard accounts, putting

those cardholders at a higher level of risk.” Pacel and Sidel (2005). (Also see Computer

Security Institute (2005) for a more detailed analysis.)

By mid-year 2005, there was a wave of security breaches and lapses that calls into

question the security of electronic financial and commercial transactions. Australian and

British press reports identified a black market in India for personal information gleaned from

financial offshore processing centers. Consumer complaints led to the arrest of employees

at a center processing Citicorp data in Pune, India. Officials at the UK National

Infrastructure Security Co-ordination Centre revealed that hackers, often linked to the Far

East, were attacking vital UK government and corporate computer networks, seeking

commercially and economically valuable information. The revelations show that computer

viruses released via the Internet increasingly are being used to garner confidential

information, ranging from personal banking details of consumers to industrial espionage.”

US investigators concurred, noting that US institutions have suffered similar attacks for “at

least a few years . . . mostly from computers in China.” (Europe WorldWatch 2005).

Some people suggest that these are simply cases reflecting the true risks of the digital

age with the implication that outsourcing and offshoring are minor additions to the mix. This

chapter argues, instead, that offshoring exacerbates existing risk and introduces new types

of risk by opening more opportunities for incursion, accident, or exposure; and it may

greatly complicate jurisdictional issues. This concern does not lead to a wholesale

condemnation and rejection of offshoring but rather to the recognition of the inadequate

attention so far paid to these risks. We hope that the issues raised here will lead to greater

awareness and thus to more prudently cautious, thoughtful, and effective practices in

preventing and dealing with these risks.

Offshoring decisions are largely business decisions and are often little influenced by

consideration of long-term risks, political consequences, or social impact. Many corporations

would argue, not unreasonably, that they do consider some long-term risks such as to

reputations, and they should and do consider risks that directly affect their business

operations, but that it is not their job to consider the social impact or possible political or

larger national security consequences of their offshoring decisions. But somebody, most

obviously government, needs to consider these impacts and consequences.

So it follows that the risks examined in this chapter come in three categories. There are

risks for companies that engage in offshoring. There are risks to individuals who are

innocent and often helpless victims of the kinds of security compromises just described.

Much of this is in the form of privacy violations or identity theft. Finally, there are risks to

the defense and economic security of nations.

Page 183

Given the subject matter and the rapidly growing number of security and privacy

violations experienced in cyberspace, this chapter is by its nature inclined to sound alarms

and encourage caution. Risks in cyberspace often have to be presented as possible or

plausible scenarios, independent of the extent to which they have occurred so far. It is

generally impossible to find accurate and comprehensive statistics on attacks and their

results, although it is clear that there is a lot going on, and many experts agree that the

problem is growing. Given the paucity of evidence, risks are discussed here in terms of the

relatively few examples that become public. Those who do not want to deal with these risks

for whatever reasons typically argue that the risks should not be taken seriously until there

is compelling evidence that the risk is real. The variety and extent of malicious activity in

cyberspace has often been underestimated in the past. Unfortunately, many of the

forecasted risks have come true. Spamming and phishing, for example, now make up a

majority of the traffic on the Internet. One dangerous risk, the use of the cyber

infrastructure to launch devastating attacks against national and international physical

infrastructure, such as transportation or electric power systems, has not yet been realized.

But this does not mean that vigilance is not needed.

Few of the risks to be considered here are unique to offshoring. But, depending on various

factors such as the laws in the countries involved, the risks may be significantly amplified by

aspects unique to the international nature of the attacks. For example, they may take the

form of exposing potential victims to a larger population of possible criminals who are not

likely to be held very accountable for the harm done to citizens of another country, or that

parts of the lengthened and expanded channels of the operation are under little or no

effective control by either the procuring or providing company or their parent countries.

Most of the information used in this chapter comes from US sources. Many experts believe

that the risk of cyber attacks is significantly under-detected and under-reported in the

United States. These problems of detecting and reporting appear to be far worse in the rest

of the world. The reasons for this are not hard to understand and probably reflect that their

citizens are not often victims of cyber-crimes, that it is difficult to find and train (and pay)

capable people to collect such information and carry out investigations, and that almost any

other form of crime probably has higher priority for the limited law enforcement resources

available in many populous, poor countries. We were also limited by an inability to obtain

and deal with locally published, non-English source material. Thus much of our coverage is

about attacks on companies and individuals in the United States. There is some justification

for this coverage. In particular, the United States offshores more work than any other

country. But it should be clear that individuals and firms around the world, not only in the

United States, are vulnerable to cyber attacks, and attackers can just as readily be located

in the United States as in some other country. Similarly, all governments that use

information technology in their critical infrastructures must face the possibility that this

technology can place their national systems and national security at risk.

6.2. Vulnerabilities: Data and Network Security and Beyond

A basic principle of security is that, the longer the supply chain and lines of

communication, the more opportunity there is to attack them. The adage that a chain is

only as strong as its weakest link often applies as the complexity of securing computer

networks is increased by routing through multiple providers. The inherent complexities in

international data communications are further compounded by jurisdictional issues

regarding regulation and legal responsibility.

Commercial or organizational alliances in the modern world rely on integrating the

computer systems of their allies or partners to some degree. Manufacturing companies

integrate suppliers into their supply chain systems. The transportation, warehousing, and

Page 184

sales systems of the distributor are linked with those of the manufacturer. These linkages

may open up additional vulnerabilities in both systems. Martin Libicki, in a forthcoming

book, identifies these as systems intimacy issues. He states, “Close relationships in

cyberspace, as in real life, can make either partner more vulnerable. A relationship, solely

by virtue of the value it brings to its partners, may be attacked by the competitors of both.

Third parties can exploit weaknesses in one to get at the other(s).” (Libicki 2006).

Outsourcing in general involves an even greater degree of intimacy because entire

business processes may be entrusted to the partner, and this often entails a greater degree

of system integration. Software development outsourcing is perhaps the most intimate

relationship of all because it constitutes a continuing impact and often access to the

procurer’s system long after the initial work is complete. Offshoring is an extension of this

intimacy across and through multiple national and international data networks under the

jurisdiction of multiple parties who may or may not be hostile to the commercial and

national interests of both the providing and procuring parties.

Offshoring risks can be categorized into systems intimacy risks and outsourcing risks.

They include the following types of vulnerabilities.

Systems Intimacy Risk

Data communications vulnerabilities. Communication channels include multiple

service providers of various nationalities. The channels are well beyond the

control of either the procuring company or the provider. Usually they are private

leased lines, which means that a certain capacity is dedicated to the buyer, but

there is no guarantee that the line is indeed private in the sense that others are

not listening. These channels are often not encrypted, or encryption is entrusted

to the control of the communications service provider. If encryption is provided,

it may not include end-to-end transaction encryption, thus leaving data exposed

at certain points along the communication path.

Loss of control over network perimeters. A link with an Offshore Development

Center (ODC) opens a broadband communications channel directly into the

procuring company that could then become dependent on the ODC for user

authentication. In one situation, security at an ODC was so notoriously lax that

its internal web servers were listed on hacker websites as useful hosts from

which to mount denial-of-service attacks. Users at that ODC were also

vulnerable to attackers hijacking their sessions to penetrate the ODC’s client

network (Ramer 2004).

Increased network complexity. Network configuration management in an

expanding and ever-changing environment challenges most IT security

capabilities. Understanding the flow of critical transactions becomes almost

impossible when an ODC is thrown into the mix. If the development center

produces software for multiple procuring companies and does not effectively

isolate the networks dedicated to each procurer, configuration management

approaches the impossible. Validating the security of trusted partners in a multiclient,

multi-vendor, mixed environment is a similarly difficult task.

Clashing security strategies. The procuring company and the ODC may take

varying approaches regarding known vulnerabilities, intrusion detection,

perimeter defense, or other security issues. These discrepancies could create

vulnerabilities for both the procuring company and the offshore provider. For

example, the procuring company could rely on very strict access control limiting

users to only those files that they need. Let us assume that the providing

company relies on very strict two-factor authentication but, once the resource

proves he (or she) is an authorized user, he is allowed relatively free range

Page 185

within the system. This situation could present a threat to the provider. A

disgruntled employee of the procuring company could access the provider’s

system and plant malicious code. Or the malcontent employee, such as someone

about to lose his job to offshoring, could use the provider’s system to launch an

attack on the procurer and thus create problems for the relationship that he

blames for the loss of his job.

Gaps in personnel security. High turnover in rapidly growing IT industries, such

as is occurring in India and the Philippines, leads to administrative stress. Even

in India, perhaps one of the better prepared of the offshoring destination

countries when it comes to security, many companies still have weak personnel

policies (NASSCOM-Evalueserve 2004). There is often a lack of personnel security

infrastructure such as searchable credit records or criminal databases (Bhat


Drastically diminished ability to know about and respond to security breaches.

Without strict and enforceable contract provisions, an offshore provider has little

incentive to notify its clients that they have had a security breach. Even if it

does, the jurisdictional and organizational issues make effective incident

response extremely difficult.

Outsourcing Risks

Loss of control over security of software development. When a company has its

software produced in an ODC, it defines the performance requirements but

relinquishes day-to-day control over software development to the overseas

vendor. Clients spend hundreds of thousands of dollars testing software

applications to ensure that they meet requirements. Rarely, however, do

security departments inspect the code for trojans (malicious software disguised

as legitimate software), viruses, or other forms of malicious code that perform

threatening or illicit activities. Virus scanners identify and sanitize widely known

viruses, but they will not find code specifically designed to sabotage or provide

particular information. Viruses are increasingly targeted at obtaining

commercially valuable information, ranging from consumer banking details to

industrial espionage (Symantec 2005). The risk of embedded malware is

enhanced by offshoring due to factors that may include less personal loyalty from

offshore contractors than from employees or onsite contractors, increased

vulnerability of the supply line, and increased potential for intervention by hostile

covert groups such as government intelligence or organized crime. Even when

inspections are possible, it may be difficult to find carefully crafted malware

hidden in large volumes of code.

Loss of control of business processes. By outsourcing to any location, a procuring

company loses a certain amount of control of the business processes that are

outsourced. There is a corresponding transfer of control over the information

necessary to perform the process. This loss of control may be exacerbated by

communications problems, cultural issues, and lines of communication that are

more vulnerable when the work is offshored. Depending on the nature and

sensitivity of the work involved (e.g., R&D or network management), this

information may be of strategic interest to competing nations and their


What seems particularly lacking within many procuring companies is an overall line of

authority and responsibility for primary data records as they pass through one, two, or more

offshore companies that perform operational tasks. Offshoring decisions are made based on

data management strategies and costs, but responsibility for security is often not

Page 186

considered. This kind of hands-off management responsibility cannot be presumed to work

in the best interests of anyone concerned with risk attenuation.

The magnitude of risk is summarized in a Symantec Corporation Internet Security Threat

Report analysis covering the period from July to December 2004. This analysis is based on

the top 50 malcode samples from output of 20,000 sensors monitoring 180 corporations

worldwide (Symantec 2004). One alarming finding was that there was a rise in threats

designed to compromise confidential information. Malicious code, including the proliferation

of trojans and bots (short for robot, a program that automatically searches the Internet for

data), created to expose confidential information or compromise systems, represented 54%

of the samples. Remotely controlled trojans and bots constituted 33% of the top 50

malware attacks, one of the most serious threats from and to offshoring. 1,403 new

vulnerabilities (more than 54 new vulnerabilities per week) were detected. Of these, 97%

were considered moderately or highly severe, meaning that successful exploitation of the

vulnerability could result in a partial or complete compromise of the targeted system.

Malware, allowing attackers to circumvent traditional perimeter security measures (e.g.,

firewalls), accounted for 48% of all vulnerabilities.

Offshoring is usually done to minimize expense, but assessments should compare total

expense for both a given level of performance and a given level of risk or protection. To

date, the comparisons have often been at the performance level without due consideration

to the risk factor.

6.3. Corporate Risks and Information Security

Corporate Outsourcing Risks

Commercial risk from offshoring is multifaceted; in today’s knowledge economy,

information security risk should be a critical issue. There are also operational business

issues including productivity, efficiency, and quality. Business managers everywhere

struggle with costs, delivery times, and product quality. Geographic and cultural spread can

adversely affect delivery times and product quality even as costs seem to be reduced.

Communication paths become longer and more convoluted; communication is more apt to

suffer distortion and error from language and cultural difference. Supply chain networks

become more diverse, less centralized, and hence less controlled. Protection from

manufacturing sabotage and theft becomes more difficult because of the size and extent of

the system. Intellectual property protection becomes more porous as infrastructure

expands on an international scale. Legal barriers and costs increase as companies cross

international boundaries, due to conflicting regulations, procedures, and practices. Safety

issues loom large, exacerbated by decentralized operational logistics.


All business depends on reliable, consistent, and clear communication. Manufacturing

processes rely on explicit process steps that companies strive to iterate and perfect.

Marketing relies on clear and concise descriptions, as well as emotional appeal. A sales

department relies on brand, trust, and perceived value. Contracts between procurer and

provider rely on all of these, along with some additional complexities. Disputes inevitably

arise, and trust is taxed; quality and deliveries will occasionally be compromised; and legal

language will be at best a palliative for a situation suddenly gone awry and not easily


The effectiveness of each of these communication attributes may be strained by physical

and legal distance and by cultural difference. Brand names in one country or language may

have an altogether different meaning, even pejorative, in another. Trust in a brand can be

damaged by local events that can have much wider ramifications. Copy exactly is a terrific

Page 187

concept for manufacturing, but if the instructions are in one language, and the operating

crew is literate in another, it may be hard to accomplish.

Sometimes the results can be catastrophic. The Union Carbide process control plant

disaster in Bhopal, India was caused by a faulty check valve that likely could have been

found by a maintenance team if they had been properly coached, but it killed twice as many

people as the September 11 terrorist attacks in the United States and wounded 100 times

as many. Many other semiconductor, chemical, pharmaceutical manufacturing, and

agricultural processing plants present such risk (Wikipedia). The damage to corporate

reputations can quickly outweigh cost advantages.

Companies with daily global interaction, for example, Boeing and Airbus, have a related

issue. Whenever a new safety finding occurs, it is imperative to reach the ground support

crews at every airport where its planes fly – which is to say, almost everywhere in the world

– as soon as possible (Flug-review). Moreover, it must be done with clear, precise,

understandable diagrams and instructions. It should not be surprising to learn that Boeing

and DuPont each publish more distinct pages of engineering text annually than any other

organization on the globe. Only the electrical engineering professional society (IEEE) rivals

them. Most companies do not have communication problems of this magnitude, but they

lack sophistication in their communication structures. Email, so often relied upon in today’s

business world, is a notoriously poor mechanism for establishing and maintaining precision.

Video and voice conferencing systems lack archival capability, focusing almost completely

on the meeting as opposed to the result. Consequently, Deloitte and Touche’s recent report

on negative experience with offshoring lists complex governance/management attention as

the leading dissatisfaction issue; this is a clear result of inadequate communication

mechanisms (Deloitte and Touche 2005).

Many people have observed that as a company grows linearly in size, its communication

paths grow geometrically. From Fred Brooks’ observations (1995) about the optimal size of

a software development team to Tom Malone’s comments (2004) about corporate

communication, it has been long established that expanded staff size and extended

geography may adversely affect communications, and therefore the effectiveness of human

transactions. Communication difficulties are not just due to offshoring; an MIT study found

that once people sit in separate buildings (even on the same site), their communication

paths seriously erode. This rule also applies to people sitting on separate floors in a

skyscraper. What is different in offshoring is that the people on the other end have much

less historic cultural alignment and, if they are working for a provider company, perhaps

much less allegiance to the procuring company’s overall mission and goals as well. Small

wonder that the previously mentioned Deloitte report (2005) cites limited transparency and

loss of knowledge among the top five issues.


Manufacturing sabotage and theft are not large issues for offshoring situations since their

costs are absorbed by the provider. They may affect deliveries, and they will certainly

affect ultimate costs, but upfront they are not particularly significant issues. On the other

hand, for offshore facilities that are part of a multinational company, sabotage and theft

have proven on occasion to be very significant. In order to understand manufacturing or

services sabotage, the context must be considered. The question needs to be considered

whether this is a problem in general that is merely exacerbated by the corporation’s size

and breadth, or whether the problem traces specifically to something inherent in the

offshoring model. For example, there are instances on record where foreign nationals

working on H1-B visas in the United States have stolen intellectual property just as there

are instances of American or European workers doing the same thing. Whether there is a

higher risk of intellectual property theft if one hires foreign nationals is an open question.

Page 188


IP issues occur at several levels. Many nations do not respect other nations’ patents or

copyrights; most require that individual patents be filed in their country. The costs of this

country-by-country protection are high; the protection afforded is variable. The most

notorious countries from a software standpoint seem to be in East Asia and Eastern Europe

(Alexandrov 2005).

Loss of knowledge is cited as the fifth most significant issue in the Deloitte report; vendor

employee turnover/training also is a high concern. These topics are broader than IP

protection since they include lore, trade secrets, and company processes. IP protection is a

risk with all outsourcing; the broader topics are more apt to become issues with offshoring

(Deloitte and Touche 2005). The entertainment industry and the software industry, both

groups whose major products are contained in codified, digitized sets of bits easily

accessed, purloined, and redistributed on the Internet, are plagued internationally by illegal

copying, sometimes referred to as software piracy.

The US Digital Millennium Copyright Act (DMCA) of 1998 is a good example of an attempt

to legislate intellectual property protection in a way that was at odds with emerging

technical capabilities. Such legal attempts to thwart the pressure of new methods are

referenced by many hackers as justification for their actions (Electronic Entertainment Policy

Initiative 2005; Gantz and Rochester 2005). Sometimes companies strike interesting

partnerships with individual countries, and sometimes countries single out companies for

sanctions. Microsoft, as the largest software vendor in the world, often has faced such

dichotomies, for example, fighting with the European Union to control source code, while, at

the same time, providing source code to the Chinese government and acceding to Chinese

rules about use by Chinese citizens and organizations in order to gain entry into the Chinese

market (Associated Press 2005).


In order to offshore work, companies face a long list of issues about international trade

including trade barriers, tariffs, taxes, import and export restrictions, currency hedges, and

transfer of partially completed assemblies versus full products, etc. The hidden costs

associated with all of these covenants and requirements can be high. The Deloitte report

(2005) focuses on this issue and singles out two topics in the top ten, namely, cost savings

questioned is sixth on the list and hidden costs is eighth.

Legal contracts consume a lot of time for the offshoring company. Terms and conditions,

notably recourse available in the event of differences, are often reported as major

difficulties requiring time-consuming, energy-sapping activities. Among the issues of

consequence is jurisdiction in the event that things go to litigation. Usually, the offshored

vendor’s country will have jurisdiction with the expected risk issues that follow as a

consequence for the purchaser and disputing party (Deloitte and Touche 2005). Some of

these business and legal issues associated with offshoring are discussed in Chapters 1 & 4.


Executive and worker exposure – personal safety – has escalated as an issue, particularly

for locales of turmoil. Hostages are taken and sometimes killed. Specific activities are

targeted for disruption: oil production in Iraq, WTO meetings in Seattle or Beijing, software

companies in Belfast, and Israeli technology companies are but a few of the targets. When

a Wall Street Journal publisher can be targeted for execution in Moscow, who can consider

himself safe? Such risks are not unknown in the United States; Charles Geschke, CEO of

Adobe, and William Hewlett’s son were taken hostage in the 1970’s in the Bay area. But a

foreign setting, especially in tumultuous areas or in areas where law enforcement

capabilities are weak, seems to raise the non-control element much higher. A particularly

noteworthy case for the IT industry was the German Red Army plans in 1986 to target the

Page 189

chief technology officers of the top 16 multi-national high-tech companies. Only one died

before the plan was thwarted by international vigilance (

Corporate Information Security Risks

Outsourcing software development or other IT-related business processes often leads to

large cost savings or quality improvements especially when the work is done in low-wage

countries such as India. At the same time, there are greatly increased risks including

financial, performance, reputation, intellectual property, and legal and regulatory

exposures. For each of these categories, it is necessary to carefully assess the risk, quantify

the potential losses, and develop cost-effective risk mitigation strategies, without which

there is no effective risk mitigation. Unfortunately, few outsourcing projects include such


The financial industry has invested heavily in risk assessment and mitigation. Banks have

spent billions of dollars on computer security to guard against fraud and theft. International

trade risks in commodities are well known, and many risk mitigation methods are in place

from payment mechanisms to insurance (a form of risk transference). Information security

risks are regularly downplayed, apparently for three reasons: (1) the failures that have

occurred are not public knowledge, (2) the exposures have been of relatively low cost to the

companies themselves, and (3) breaches are less tangible than, for instance, ships sinking

at sea, physical bank robberies, or highway accidents. Some Indian IT outsourcing service

providers have been more publicly concerned about information security than the Western

companies procuring their services. Numerous leaders of the Indian IT industry have related

that they are concerned about security, but, as business managers, they probably will not

invest more in security than is required by their clients (NASSCOM; Ramer 2002-2003).

Procuring companies have been conspicuously quiet about security. This apparently

curious fact can perhaps be explained by realizing that, when people perceive a security

threat, they act to avoid it or protect themselves against it, but if they do not perceive the

threat, they do not worry about it. Procuring companies also downplay information security

to avoid the threat of negative public opinion and potential regulation. Both of these

responses, while rational in some ways, do not often proceed from a concrete analysis of

the actual risks involved in the projects. In contrast, leading Indian provider companies

have identified client security concerns as an obstacle to growth and, through the Indian

software trade association, NASSCOM, have initiated campaigns to enhance security

awareness and change perceptions.


Risks turn into incidents through two basic kinds of action, accidents and intentional acts.

Many are direct, for instance, a computer system fails on-site, a disgruntled employee

sabotages the equipment, a well-intentioned employee makes an error, or an external

hacker perpetrates an effective denial-of-service attack. There can also be collateral

damage where an external incident or accident causes incursion. While accidents can lead to

significant damage, this discussion concerns threat actors. These incidents are arguably the

most dangerous of anticipatable incidents because they are carefully targeted. Controls

against threat actors also help guard against accidents as well. Accidents caused by human

error or acts of nature are an essential part of disaster recovery and business continuity

planning that are not in our scope.

We briefly discuss six such threats: rogue employees, hackers, organized crime

syndicates, industrial espionage, unfriendly nations, and terrorists.

Rogue employees. Citibank customer complaints of fraud led to the arrests in

April 2005 of former employees of a call center in Pune, India. They were

charged with defrauding Citibank account holders of $300,000. In this case, four

Page 190

rogue employees accessed account numbers and PIN numbers to transfer money

out of the accounts (Computerworld 2005). Rogue programmers have installed

back doors into code, trojan programs that send out sensitive information, and

logic bombs that sabotage operations.

Hackers. The term refers to a special breed of programmer characterized as a

person who is simply intellectually curious without evil or financial motivations;

increasingly it has morphed to include individuals who may have dark intentions,

varying from a self-described desire to thwart e-commerce, in general (e.g., viral

attacks), to targeting specific companies for specific vendetta reasons (Electronic

Entertainment Policy Initiative 2005).

Organized criminal syndicates. Criminal syndicates around the world regularly

engage in identity theft for financial gain. Bruce Schneier, CTO of the security

firm Counterpane, has said that there is regular trade in credit card numbers and

the only reason that most of us have not experienced fraud is that the thieves

have not yet had a chance to use our account number (Schneier 2005). In May

2005, a criminal syndicate in New Delhi sold access information to 1,000 British

bank accounts. The information was collected from a network of employees at a

call center that the British banks had outsourced to. (The Hindu 2005).

Industrial espionage. Competitive intelligence and industrial espionage are

supposedly separated by an ethical wall and a legal structure. But in countries

where information theft is not illegal, the dividing line evaporates. Offshoring

increases the possibilities and profitability of these activities, while decreasing the


Unfriendly nation states. More than 30 states have been identified as developing

cyber warfare capabilities. A number of these techniques have been extensively

studied for impact. Offshoring to states that have lasting conflicts of interest with

the home state of the procurer, whether in legal jurisdictions or other disputed

matters, heightens risk elements (Billo and Chang 2004).

Terrorists dedicated to attacking national interests. Numerous organizations are

dedicated to cyber attacks on Indian IT sites. Groups of Pakistan-based cyberhackers

have routinely defaced websites and claimed to have penetrated the

perimeters of Indian IT companies. E-jihad sites have launched extensive denialof-

service attacks on US, Israeli, and Indian targets (Institute for Security

Studies 2003). Indian outsourcing sites came under attack by jihadi groups in

Bangalore in June 2005 and in Mumbai in 2003.

Corporate Strategic Risks of Outsourcing

Outsourcing key business functions can create a strategic risk that is often disregarded in

the day-to-day drive to cut costs and meet deadlines. In the modern hyper-competitive

culture of international business, alliances are critical because markets do not allow time for

companies to respond to competitive challenges by developing their own capabilities.

Instead, a new product or competitive advantage is more easily challenged by entering into

an alliance with an existing company that already has that product or capability. However,

as Martin Libicki, a security researcher at the RAND Corporation, points out:

“Third parties may want to attack a relationship simply because it is the heart of an

opposing alliance. Sundering relationships can render opposing alliances less


The logic of sundering mirrors the logic of binding. The ability to form

coalitions … is of growing value in competitive arenas. Coalitions, these days,

float on the exchange of information; notably the privileged exchange of

Page 191

sensitive information (much as personal relationships are ratified and

maintained through the exchange of favors) such as inventory data (e.g.

Proctor & Gamble’s evolving relationship with Wal-mart) or design information

(e.g. for new cars). The greater the importance of proprietary and personal

information flowing among enterprises, the more important is the ability to

protect such information to its cohesion. Thus the more important good

security is to the choice of partners.” (Libicki 2006, 9, 3)

There is another set of potential problems associated with outsourcing and that is

exacerbated by offshoring. These problems relate to the fact that virtually all procurerprovider

relations in the domain of interest to us involve connecting or sharing the

information systems of the procurer and the provider. This happens to greater or lesser

degrees with greater or lesser vulnerabilities, depending on how it is done.

Various security vulnerabilities can result from this relationship. For example, the procurer

allows extraordinary and unsecured access to the provider, the procurer may even have the

provider at least partially provide security to the procurer's system, or the procurer could

become more dependent on the provider. One can imagine how the two systems can be

fairly secure in different ways and how connecting them could create a joined system that is

less secure than either one. The connection could expose the procurer to security problems

from the provider or expose both to security problems from third parties. Both procurer and

provider companies must conduct risk assessments to make sure that the explicit ways the

systems get connected do not open such vulnerabilities

6.4. Risks to the Individual: Privacy and Identity Theft

A contentious and challenging aspect of offshoring is its risk impact on individuals.

Individuals are pawns in many respects in this global restructuring of business, but they

stand to bear the brunt of many issues as risks occur: loss of privacy, loss of jobs, loss of

property, and loss of security are and will continue to be experienced at the individual level.

Some facets affect employees, while other exposures impact customers. Many effects will

be borne by the general population within the home country of the procuring companies;

some effects will impact the citizens of many countries. Regrettably, for the most part,

individuals will have little to say or do to protect themselves.

Section 6.2 dealt with data security. Without data incursion, there is seldom an issue so

protection against the risk of data intrusion is the first order of business. Businesses,

governments, and military groups understand a wide range of issues pertinent to data

security, and they can make decisions and put policies and procedures in place to mitigate

risks. Individuals, though, have little impact on data security procedures or policies.

It is therefore fundamental to describe the risks and exposures of offshoring from the

point of view of the individual and to suggest some possible mitigation strategies for them.

First, note that so much offshoring has already occurred that the risks are in place and must

be dealt with. This topic is politically charged. Many people, particularly in the United States

and Europe, have an increasing feeling that THEY are putting OUR jobs, financial records,

health records, and privacy at risk. Some people believe that national security is being

compromised as well. Such views, to the extent that they become salient, can have

significant political impact (Knox 2005).

What are Privacy Rights?

An offshoring issue of great consequence is the differing cultural and legal definitions of

privacy around the world. Personal data of tens of millions of individuals are widely

available. Individuals who would make illicit use of this data may have vastly different

Page 192

geopolitical, cultural, and legal environments than those whose private data is being used.

The goal may be criminal as perceived by the victim and his home country, but not

necessarily illegal or punishable from the point of view of the extant government, court, or

culture in which the perpetrator lives. Historically, citizens have looked to their own

government and its legal system for the protections to which they believe they are entitled.

Consider the issues raised if data about AIDS patients is purloined by an extortion group

in a web-based cell in the provider country, followed by a set of threats to expose the

diagnoses to employers, insurers, and neighbors. Privacy has certainly been invaded,

financial impact could be severe, and the social cost to the individual is incalculable. Where

could victims of this kind of action turn? The fact is that provider countries often have a

very different set of laws regarding citizen rights than do the procurers. Those laws are

interpreted or enforced with respect to theft of data on local grounds rather than with

reciprocal rules. Thus, some nations’ laws invite certain behaviors that other countries

would consider illegal.

Privacy in Some Leading Procuring and Providing Countries

A core function of any nation is to secure its borders, including those less tangible and

porous boundaries of the information space. Not all nations seek to protect the privacy of

their citizenry. In the United States, when privacy conflicts with free speech, the right of the

speaker rather than the subject dominates. In Europe, privacy is protected assiduously by

most nations, but even there, free speech is also encouraged in ways that abet privacy

assaults on occasion. Other countries, such as Israel, have strong laws governing privacy

around medical and other sensitive personal areas, but their governments also have

histories of dealing strictly with perceived threats to the public welfare in ways that may

trample individual privacy rights on occasion. China, which for years has been a concern of

human rights groups, presumes that the Four Cardinal Principles govern the nation:

Leadership of the Chinese Communist Party, Marxism-Leninism-Maoism Thought, People's

Democratic Dictatorship, and Remaining on the Socialist Road. These principles are not

supportive of individual privacy rights.

European companies in the main adhere to strong data privacy protection, ensured by

zealous data auditing and control. Swiss banking data privacy is legendary, but privacy of

individual data – especially health and financial data – also has a long history of protection

across the continent. By contrast, India, China, and the United States have been much

more open with personal data; selling consumer lists to advertisers is a good example of a

common practice in the United States that could be considered infringement upon an

individual’s privacy. Because of relatively unique European history, many European

countries have been loath to send data to countries where the data is not strongly

protected. India has proposed information privacy policies for offshoring company data that

square with European data privacy policies, and these policies could provide a potential

competitive advantage over the United States for offshoring work originating in the

European Union (Peterson 2002).

There is no single directional arrow in terms of privacy and offshoring. In some cases,

data are offshored to areas where there is stronger protection, and, in some cases,

offshoring creates privacy risk distinct from security and operational risks. In order to

examine the general issue of privacy more closely, this section will consider several major

locations of offshoring and then discuss a set of possible responses.

The European Union

Europe and Canada have the most comprehensive data protection systems of any

countries. Countries that send data to Europe can expect protection equivalent to that of the

sending nation because the European Union (EU) harmonized and coordinated twenty-five

national regulatory systems under the Data Protection Directive in 1995. Each nation

Page 193

develops its own implementing legislation that complies with the Directive. The Directive

was issued both to establish minimum standards on the fundamental right of privacy and to

ensure the free flow of personal information between states.

The European Data Protection Directive restricts the data that may be compiled, and it

controls data once compiled. There are data that may not be compiled if privacy violations

would create human rights violations: sexual orientation, religion, and racial identification.

These requirements come from the basis of privacy as a human right, that is, privacy as a

right of autonomy. The substantive principles underlying the directive are that data must

be purpose-specific in collection and processing, relevant to the reason for processing,

accurate, and deleted when the stated purpose has ended. There must be unambiguous

consent by the data subject for data collection.

Substantive consent requires notification by the data controller of the identity of the

controller and the intended uses of the data. Other information that must be provided

before data collection occurs includes the consequences of not providing data, rights of

access and correction, and any exceptions for research. The rights of access and correction

ensure data integrity by ensuring that subjects can correct, erase, or block inaccurate data.

After the directive was developed in 1995 and implemented by 1998 in most member

states, a concern about the lack of data protection in the United States became urgent.

Data flow could not simply continue unconstrained to the United States since, from the

perspective of the European Union, the United States is an unregulated data haven. The

European Union strongly encouraged the United States to harmonize its own privacy

regulation with the directive; however, the request was rejected by both the US executive

branch and the Congress.

In the 2000 Agreement on Safe Harbor Principles, the European Union and the United

States developed a process to prevent an interruption of the data flow from Europe to the

United States. The Safe Harbor Act requires American companies to develop privacy policies

that align with the Data Protection Directive, inform European customers of their privacy

rights under this policy, create easy-to-use complaint mechanisms, register with

independent dispute-resolution mechanisms to resolve complaints, and notify customers of

any change in policy. However, few companies have signed up for Safe Harbor. At the first

anniversary of the Safe Harbor, only 54 companies had registered and complied with the

Safe Harbor guidelines (Peterson 2002).

In 2001, the Data Commission approved standard contractual clauses for Data Transfers

to Non-EU countries to deal with those nations where neither implementing legislation nor

Safe Harbor agreements exist. The Safe Harbor gives EU citizens protection and compliant

American companies a relative advantage in obtaining offshoring contracts from EU states.

Despite the apparent value in terms of low cost, strategic advantage, and protection from

liability, few American firms have taken advantage of the Safe Harbor. Thus, at this point

European companies have only a few low-risk choices for offshoring to American firms.

Most American and Indian firms to which EU companies offshore must make custom data

protection agreements.

The United States

From the perspective of the European Union, the United States is an unregulated data

haven. In the United States, personally identifiable data can be accessed by those who

assert a legitimate business need even in business sectors where privacy protection exists,

for example, finance, health care, and telecommunications. Sectored legislation includes

the Driver’s Privacy Protection Act, the Video Protection Act, Electronic Communications

Privacy Act, and elements of the Health Insurance Portability and Accountability Act (HIPAA,

1996) (Swire and Steinfeld). Similarly, the Gramm-Leach-Bliley Act (1999) was designed to

increase individual financial privacy (Janger and Schwartz 2002).

Page 194

However, the US approach has proven inadequate. Authorities such as Bruce Schneier

have criticized the accuracy and integrity of personal data aggregated by commercial

brokers (Schneier 2005b). A recent and widely scrutinized example makes the limits of the

US approach clear. ChoicePoint, a commercial data broker, was created as a separately

owned subsidiary of Equifax, Inc., and a copy of all Equifax data was transferred to the new

company (Solove and Hofnagel 2005). As a subsidiary, ChoicePoint was not required to

comply with the privacy regulations governing US financial and credit institutions.

Consequently, the regulations that prohibit Equifax from selling its data do not apply to

ChoicePoint. Indeed, there have been several news reports in recent years alleging

ChoicePoint's disregard for accuracy in its data, and it is this exposure of personally

identifiable (and often inaccurate) data that places individuals at risk for identity theft.

ChoicePoint acted explicitly to purchase the Mexican voter rolls which are protected under

federal Mexican law (Peralte and Ferris 2003). The three Mexican nationals who sold the

data to ChoicePoint were prosecuted, but ChoicePoint itself was not subject to Mexican

federal law and it still markets the data internationally. ChoicePoint was also subject to

scrutiny in 2000 when the listing of Florida felons provided to purge the data rolls was found

to have systematic factual biases against African Americans (Pierra 2001). The basic

elements of data protection – notification, consent, auditing, and accuracy – are all absent

in ChoicePoint processes (Solove and Hoofnagel 2005). The ability to commit a felony in one

nation (Mexico) and then use the results of the felonious data collection, illustrates the

limits of the reach of national laws in a networked global economy. The inability to identify

a basic classification (living in Florida, convicted of a felony) reflects the risk of lack of data

integrity when there are no data protection requirements. In mid-2005, the company

announced major changes in its policy and approach. How wide-ranging the results will be

remains to be seen (Wall Street Journal 2005).

In the United States the approach to privacy has been specific to particular sectors of the

economy. When abuses of data in a business sector are identified, they are addressed by

legislation directed to that sector. In protecting data, the conflict between autonomy and

seclusion is often implemented as a distinction between prohibited data (no one can ask for

the data); opt-in (you must be asked for the data); and opt-out (you must pursue the

opportunity not to be included). For example, the Gramm-Leach-Bliley Act allows customers

to opt-out of data marketing.


Privacy as autonomy underlies many of the sectored laws that have been implemented in

the United States. Privacy in the choice of health care for women, US postal mail, and

personal memberships are all grounded in the right to autonomy. The essential observation

of privacy as autonomy is that people under surveillance are not free. Actions taken with

knowledge of direct data surveillance will be more constrained than actions taken

anonymously. While this is sometimes taken to mean that anonymity is not accountability,

the freedom of citizens to interact with government and the anonymity necessary for

whistle-blowers illustrate the false dichotomy.

Without anonymity for those with little power, those with power have lessened

accountability. Anonymity in a democracy is a critical factor in accountability. Perhaps the

classic example is the secret ballot for the voting process. This was a hard-won position in

the late nineteenth century that today is seen as an inalienable right. The right of

autonomy was first defined in the United States legal system by Justices Earl Warren and

Louis Brandeis, who also coined the famous phrase right to be let alone that underlies

privacy as seclusion.

Page 195


While the right to be let alone, the right of seclusion, seems at first irrelevant to the issue

of offshoring, the changing mores of the web and some of its abuses have brought this

segment into sharper focus. The most prevalent violation of seclusion privacy is spam,

which frequently originates offshore. In the United States, the CAN Spam Act was seen by

many in the anti-spam community as legalizing spam, and by some in the ISP community

as providing a uniform legal mechanism for prosecution of spammers. While CAN Spam

created a national law, it overrode many stronger state anti-spam laws (Ford 2005). In

typical American fashion, specific legislation was enacted to deal with direct abuses, for

example, the Do Not Call list and the opt-out provision in Gramm-Leach-Bliley.


The United States is a nation with strong respect for individual property rights, although

these rights are constantly being tested. The right to property extends to personal data as

property. Excluding specific sectored protections, data in the United States are currently

regarded as property. Subject rights over data are lost when data are disclosed because

property is an alienable right. That is, once sold, there is no longer a personal interest in

property any more than one retains a legal right over a house after it has been sold. The

property interest in data then becomes entirely the interest of the data owner which is how

a data broker is empowered to operate. Thus, the data broker has no direct customer

relationship with the subject of the data, and consequently the broker owes the data subject

no duty of care. Similar lack of care governs most data considered to be in the public

domain. American law even allows for public sector data to be priced and delivered through

aggregators and data brokers.


Data protection legislation has much in common with other privacy legislation: notice,

consent, integrity, and exercise of rights are all germane. This commonality is based on the

Fair Information Practice Principles (Privacy Protection Study Commission 1977). Notice

requires that no compilation be secret. Consent requires that data be used only with the

consent of the subject. There is continuing contention between passive consent (i.e., optout)

and requirements for active consent (i.e., opt-in). Integrity requires that data are

correct. The most common citizen interaction with integrity is with credit records. Credit

reporting organization must provide citizens with credit reports and the right to redress if

data are incorrect. Before the 1970 Fair Credit Reporting Act, credit data were often based

on gossip (e.g., investigative credit reports) and other spurious sources. Errors in credit

reports could not be corrected by the individuals who were the targets of the reports.

Similar errors are found today in reports on individuals prepared by data brokers (Schneier

2005b). Redress, access, and enforcement are the mechanisms by which data subjects are

ensured integrity. Access requires that individuals are able to view data about themselves.

Redress requires that there are mechanisms to correct data.


India’s property model is analogous to that of the United States, allowing authorized use

of personal data. Since the 2000 Information Technology Act, strong prohibitions have

been in place regarding data theft (Government of India 2000). India is a bastion for

freedom of speech and autonomous action. There is no comprehensive governmentfiltering

regime for Internet content. Internet kiosks flourish with resounding condemnations

of government policies, and different social groups engage in battles of words without

governmental oversight or intrusion. Anonymous posting is allowed. For an Indian youth,

the cultural environment is fully conducive to viewing the web as a simple extension of the

local mores. Hence, dissent and spamming about one’s beliefs targeting another country not

only go unsanctioned, but may be endorsed or supported culturally.

Page 196

On the other hand, India has a significant regulatory structure in medicine,

telecommunications, elections, and other industries deemed critical by the government. For

example, election equipment is designed not to show public totals, and votes are held to be

private. Each of these regulatory or surveillance regimes concerns itself with an individual’s

privacy. For example, when the Telecom Regulatory Authority received a proposal by the

major carriers in India for publication of a cellular phone directory, the outcry could have

been lifted from any Western paper. Health care studies examine the physical privacy

provided for patients in terms of examinations and discussion of diagnosis under the

medical regulatory authority. Thus while there is not specific privacy regulation in distinct

sectors in India, the regulatory bodies in medicine, telecommunications, and elections

address privacy rights as part of their regulatory function. As a result of India’s

involvement in long-term struggles against terrorism in Kashmir and elsewhere, wiretapping

and searching public discussion areas for activity in regard to terrorism are not


Data are primarily alienable property in India as in the United States. Data are private

property so governmental seizure would have to be strongly justified and public in nature.

There is no Safe Harbor agreement between India and the European Union. The current

intellectual property regime, and lack of corresponding enforcement, suggests that an

enforcement regime on data as property might not be effective.


China has a communist government and an institutionalized ruling political party, thus the

very concept of personal privacy is contrary to the underlying philosophical organization.

China is organized on the Four Cardinal Principles, listed earlier in this chapter, none of

which have implications that extend protection to individual privacy. Criticism or violation of

the Four Cardinal Principles is prohibited. The overall data philosophy is concerned with

state control of Internet use rather than citizen privacy. For example, Internet-based

discourse in China has a series of disapproved words and a prohibition of criticism of current

leaders and the Four Cardinal Principles. Anonymous electronic speech is officially prohibited

in China. Speech on the Internet in China is controlled through technical means (filtering of

postings, prohibition of websites, and detection of encryption) and political means

(punishment of those who receive or transmit unacceptable words, ideas, concepts, or

content). Microsoft has recently acceded to the Chinese government on this point with

respect to blog management.

In some cultures, privacy includes the right to seclusion or freedom from excessive

intrusion. For example, the right to seclusion in Britain includes the right of citizens to ask

to be excluded from junk mail where, as in the United States, this only extends to spam and

marketing phone calls.. However, there is no right to seclusion in the People’s Republic of

China. Spam is an approved and active business in China. Asia as a region has had such

significant problems with spam that many IT operations blacklist all Chinese incoming mail.

For example, both China and Korea have been one-click selections for blocking on Spamcop

for years. China is active in hosting websites for spammers and supporting the market for

lists for unsolicited bulk email. For example, the Hong Kong-based offers, in

English, 238 million email addresses for spamming for any business. The information

includes more than 10 million AOL, Hotmail, Yahoo, EarthLink, and MSN addresses as well

as 1 million personal profiles complete with name, address, email, birthday and country,

presumably harvested from the profiles of those advertisers. The target of the web page,

the bulk of the email addresses, and the cost of removing that email create costs that are

borne by individuals outside of China, yet the laws of their home country pertaining to spam

do not apply. Recent research shows that free email accounts receive an order of

magnitude more spam than legitimate email from China (Hulten, Goodman, Rounthwaite


Page 197

Identity Theft and Credit Card Fraud

Many people are concerned about the ease with which their identity or their credit cards

can be stolen. Identity theft can lead to property loss (commonly thought to be only bank

account theft, but it can involve property deed transfers or transfer of income allocations)

and damaged credit worthiness, and it often involves a long hassle with a multitude of

faceless and possibly irresponsible organizations in order to clear a besmirched credit

record. It can even lead to a person being labeled a fugitive felon and cause their lose

voting rights to be lost when a stolen identity is used in committing a crime.

In the United States where identity theft appears particularly acute, a major factor is that

companies are able to sell Social Security numbers tied to names of individuals, complete

with addresses, birth dates, and other pertinent information that enables not just invasion

of privacy, but also the alteration and use of the data for criminal intent.

Identity theft and credit card fraud are huge problems globally. Given the wave of

incidents within the United States in 2005, as described at the beginning of this chapter, it

may not be saying much to suggest that consumer data are at any greater risk of exposure

in an outsourcing provider country such as India rather than in a procuring country. But it

could be more likely that the events will be brought to light and somehow constructively

dealt with in the country where most of the victims are citizens. However, privacy is a key

issue in the debate over offshoring. Business Week Asia ran a story on this in August 2004:

“186 bills that aim to limit offshore outsourcing are pending in the U.S. Congress and

40 state legislatures. Dozens of those involve restrictions on transmission of data.

For example, the SAFE ID Act, sponsored by Senator Hillary Clinton (D-N.Y.), and a

similar House bill by Representative Edward J. Markey (D-Mass.), would require

businesses to notify U.S. consumers before sending personal information overseas --

and would bar companies from denying service or charging a higher price if

customers balk. Although no such bills have been enacted so far, "next year I think

all of this legislation will be back and spike up again as a huge issue, especially if the

U.S. recovery stalls”, says R. Bruce Josten, US Chamber of Commerce, who helped

industry fight the legislation.” (Engardio 2004)

The particularity of identity theft in the United States is compounded by the fact that the

criminal liability and recourse when an American is defrauded is far from clear domestically

and is further complicated by offshoring. Theoretically, the US company that farmed out the

work is legally responsible. Indian call centers usually sign their contracts in the United

States. Thus, both offshoring procurer and provider can be sued in domestic courts by their

corporate customers. However, liability for international security and privacy data breaches

is unsettled in case law. Americans must often begin with local police to make a claim of

fraud. Many local police departments lack the personnel to address individual fraud cases

and are ill suited to address complex international technical jurisdictional issues. A problem

with privacy risk in the United States is the likelihood that the large organizations that are

most able to mitigate risk instead transfer the costs to individuals who are left without

jurisdictional recourse especially when the data is offshored.

A number of cases have surfaced, including the situation at MphasiS, one of India’s

largest call center providers, where Citibank accounts were penetrated and the events were

only found by account holders who called the bank to complain. Nonetheless, an Indian

official with MphasiS said later that week:

"While we are unhappy with the incident itself, we are at the same time quite

pleased that detection systems worked. While such incidents unfortunately do

happen everywhere, timely and exemplary enforcement ensures that no-one needs

fear that culprits or potential culprits can get away and the reputation and credibility

of the entire system is actually preserved and enhanced." (McCue 2005)

Page 198

One long-term American consultant to India, when asked to comment on how well the

enforcement provisions really work, was quite candid:

“There is no way that the company itself will be prosecuted. MphasiS is one of the

top ten providers and their President is the current president of NASSCOM. The

individual perpetrators will be prosecuted under a Government of India act but as is

typical with Indian justice, it may be years before it comes to trial. For example,

there was a bribery scandal in the 1980’s involving Bofors, the Swedish defense

supplier, and before the case came to trial in the late 1990’s several of the accused

had died of old age. NASSCOM advocates strong security to its members but it

doesn't really have any enforcement power. The only enforcement provisions that

would really be effective . . . would be pushed from the demand side. In my work

over there, I heard again and again, that providers will conform to whatever security

measures the customers require, with an implied ‘but unless they require them we

will do the minimum we deem necessary’. Therefore, if US companies aren’t acting

to protect their clients then the government has to step in and protect the privacy of

its citizens." (Ramer 2005)

Not surprisingly, most companies in offshoring businesses assert that these were isolated

instances. Other observers are not so sanguine. The day after the MphasiS story broke,

TBR News had a follow-up story about 310,000 accounts that were illegally accessed via

Lexus/Nexus, coupled with the 145,000 that had been fraudulently exposed at ChoicePoint.

These predated the 40,000,000 ostensibly accessed records reported two months later at

CardSystems. While offshoring did not figure in all of the reported breaches, the net effect

has been unsettling for the data handling industry (Timmons 2005; Rigby and Kolker 2005).

Following the April 2005 incident at MphasiS, many in the Indian IT industry called for

serious reforms and security improvements, including calls for a law on data protection as

well as more stringent laws on enforcement of contracts (Thiagarajan 2005). However, two

months after the MphasiS fraud case, a new scandal broke out. A reporter from The Sun of

London was able to purchase the account information of 1,000 British bank customers for a

price of $5,000. The reporter was told the information came from a network of Delhi call

center workers. The contact person boasted that he could provide information on 2,000

accounts a month (Harvey 2005).

Dealing with the Risks and Exposures for Individuals

India’s national leadership is seeking data protection legislation to directly compete with

US firms for work involving data that are currently offshored from Europe to the United

States. Such a change in the competitive landscape would increase the challenges of

globalization for the United States in terms of long-term learning, productivity, and

employment. The United States has a structural decision to make with respect to

international competitive strategy, that is, whether (1) to compete as a high-quality service

provider with security and privacy used as competitive advantages, or (2) to compete in the

global market as a data haven.

As more and more countries get into the offshoring game, the price pressures on

providers of offshored services only increases. According to banking industry sources,

effectively securing a transaction can add 15 to 18 percent to the cost (O’Bryan 2003). The

real costs of offshoring should include legal, security, auditing, and contingency planning

costs, all of which increase when offshoring. A critical but often overlooked issue is that

many offshore providers do not perform realistic annual disaster recovery testing. Instead

they test with a limited number of client companies at a time. With increased price

pressures, the temptation to skimp on security measures strengthens. Thus the need for

common and verifiable security standards gets stronger as well.

Page 199

Outsourced IT-enabled services, whether the service is software development or loan

processing or even a call center, involves interaction between the procurer's network which

probably is more controlled, known, and trusted, and another network which the client has

much less control over, less knowledge of, and can trust less. Networked security will be

greatly enhanced if verifiable security standards for offshoring are put in place.

Politicians in a number of US states, as well as in Congress, have begun inquiry into some

of these risks and exposures. Of the thirty-six states in 2004 that sponsored legislation that

would limit offshoring, only two enacted bills that year. But in 2005, sixteen states had bills

introduced in the opening sessions, a clear indication that concern has not abated (Cooney

2005). Given the recent spate of high-profile cases and high number of affected people,

much more legislative involvement can be anticipated. This is worth putting into some

longer-term perspective. Just as other technologies advanced and enriched the early risktakers

and owners, there were frequently undesired consequences. Pollution from power

plants is but one of dozens of examples one could cite. When the undesired consequences

rise to impact a sufficient number of citizen’s rights (clean air, water, noise), governments

generally rise to the occasion to pass laws to protect the citizens. Looking back, the lag

time is generally substantial and great harm may occur before citizens acting through their

combined power of government set out to seek remedies.

We are now a decade or so into the use of widespread computer networking in which

individuals can be brought into harm’s way with little to no financial risk to those who

actively or negligently inflict harm on others. Just as with sprinklers and fire codes, speed

limits and air bags, clean air regulations and smoke stack scrubbers, both legal and

technical means can play a role in the information sphere to protect the rights and assets of


The problem of vetting offshore providers in today's world is complicated. Procuring

companies are primarily focused on obtaining the financial benefits of offshoring; most

appear to be naive about the risks, or they do not have the time or resources to care. This

may account for why this kind of topic so seldom arises in discussions of offshoring.

Government vetting in today's world (e.g., by greatly extending the scope and authority of

the activities of the interagency Committee on Foreign Investment in the United States)

would likely be a nightmare. The best defense is a set of policies that protect against giving

providers strong forms of access/control which, in addition to raising the security/privacy

concerns, can also make the procurer dependent on, and perhaps hostage to, the provider if

it is done foolishly. (Contrary to popular opinion, export controls were remarkably effective

for most of their existence for several complex and reinforcing reasons. Industry has often

argued against them on the basis that they stifled technological progress, that technology

flows were impossible to retard, etc., but those concerns really didn't come into play until

the late 1980s.) What can be done in individual jurisdictions is to prevent the transfer of

risks to individuals who are the least able to mitigate or recover. Entities that choose a risk

should be the ones who pay or profit from the risk premium and any downside. The problem

of weak or more seriously compromised provider organizations is considered further in

Section 6.5.

It would be highly desirable if economic incentives and competition would be sufficient for

companies, both the procurers and providers, to effectively protect the privacy of their

customers. So far, as is the case with regard to other forms of cyber security problems, this

has not proven to be the case, and it does not seem to be improving as rapidly as the

increasing numbers of violations and victims. More generally, the following measures might

be considered by lawmakers or regulators and could also be included in offshoring contracts

for dealing with situations where there are risks of privacy and identity theft.

Page 200

For provider companies:

Providers should have security and data protection plans. They should be

required by contract, and work should not be allowed to begin without them.

There should be clear requirements for reporting incidents. Breach should be

grounds for termination and financial redress

Providers should be certified in some way, perhaps through adherence to

prescribed standards. The risk is that such standards only provide cover for

malfeasance and not true protection. A difficult question is: Who would certify the

providers and effectively stand behind the certifications? It would clearly have to

include government parties in the provider’s home country.

Offshore providers should agree to no indirect third-party outsourcing without

explicit approval from the procurer. This should be contractual, with high

sanctions, for example, grounds for termination.

For provider countries:

Provider countries should enact data privacy laws that apply to foreign citizens

whose sensitive data is offshored to their country, or agree to recognize the laws

of the procuring countries as applying to foreign citizens and make them

enforceable in the providing country. Violation of national privacy laws, in

addition to breaches of contract, should be covered.

These laws should be backed by either demonstrated capacity to enforce (e.g.,

by a good record of enforcement) or by secured assets in order to ensure penalty

Providing countries should be certified as Safe Harbors as is done by the

European Union, but in the more general context of the procuring country and

the foreign citizens who are vulnerable to the compromise of their sensitive data.

For procurer companies:

There should be reporting requirements and stiff fines for failing to protect

sensitive information – just like failing health inspections, speeding, or polluting.

While a procurer may be theoretically subject to privacy regulations, experience

shows that practice is woefully lacking (Ramer 2005b).

For procurer countries:

They should consider legislation or other strong forms of regulation requiring any

of the measures listed here.

Certain kinds of information about a nation’s citizens or businesses may be

considered to be particularly sensitive and vulnerable. Consideration should be

given to reviewing such categories and banning certain data from being hosted

outside of the originating country.

Technical means:

There should be no mass export of databases or transactions. Databases should

be kept on servers in the procuring countries. This would also make it easier to

cut off a derelict or abusive provider.

Data should be used in transactions on a one-record-at-a-time and as needed

basis. After one transaction is completed, another should not be initiated until the

record for the first is effectively removed from access.

Page 201

Databases should be encrypted to help protect data at rest and in transit and

prevent unauthorized data mining for purposes not intended by the procuring

organization or contrary to relevant laws.

Systems should be instrumented to facilitate incident discovery, reporting, and


6.5 Risks for National Capabilities and National Sovereignty

One important aspect of offshoring risks that is often ignored or treated perfunctorily is

the impact on national capabilities. Sovereignty is basic to a national government’s reason

for existence, and effective sovereignty must include national defense, national economy,

and national well-being. While individuals and even companies may bear the immediate

and visible brunt of IT globalization, including loss of jobs, compromise of data, and loss of

intellectual capital, the overall social impact must be evaluated for a full contextual

understanding of the impact of offshoring. In this regard, the IT issues that are addressed in

this entire report have considerable consequence to the national interests for many

countries. Thus this section examines the threats to a nation’s sovereignty that are

exacerbated or introduced by IT outsourcing.

Effective sovereignty must include a national economy that is able to provide for its

citizen’s well being and is not subject to arbitrary manipulation by external forces. As

economies have moved from bricks and mortar, and rail and road infrastructures, to an

information technology-controlled infrastructure, offshoring of IT raises two key risks,

namely, the vulnerability of infrastructure or defense systems to remote electronic attack,

and the loss of the ability to fix or replace economic infrastructure

Modern economic infrastructure is dependent on an increasingly global IT network and

vulnerable to remote attack through inter-networked systems. Operators in Mumbai or

Manila help customers with credit card transactions. Programmers in Bangalore or the

Ukraine maintain computer operations for European airlines. Railroads, power companies,

and defense contractors regularly use global outsourcing to cut costs and deliver services to

their clients. Unauthorized hacker access to these systems could, with malicious intent,

cause blackouts, air or rail accidents, or communication system shutdowns. Financial fraud

or sabotage of financial system through cyber attacks could be more devastating than

physical attacks.

Modern defense systems are arguably even more dependent on information technology

than infrastructure systems. From fighter aircraft to command-and-control systems to

robot-bomb detonators, software is an essential ingredient. Defense systems developed

without proper controls significantly increase the risk of weapons systems failure or

sabotage. It is in this context that cyber warfare presents a serious threat.

Earlier sections of this chapter identified the mechanisms by which offshoring increases

the risks of IT systems development and maintenance. It expands the range of process

vulnerabilities and widens the field of potential threats, thus offshoring significantly

increases the risk of a successful infrastructure attack or the compromising of weapons

systems. In such situations, offshoring can undercut the national capability to repair and

replace these critical components of a nation’s defense.

Information Technology is critical because it is intimately bound up with technological

innovation. The ability to take an engineering advance and create a functioning software

system is a critical part of the process of technical innovation. Therefore the future

economic welfare of a nation can be put at risk if it is unable to reproduce technological

innovation at a sufficient rate to remain competitive with other players. Concern over

Page 202

investment in innovation in developed nations, particularly the United States, is discussed in

Chapter 8. The impact of offshoring on infrastructure, defense systems, and national

capabilities to remain competitive will be discussed in the rest of this section.

Rising Threats to Infrastructure and Military System

Commercial Off-The-Shelf (COTS) product purchasing strategies have been adopted by

the United States and other countries in building their IT-based military systems. These

countries have also shared national and international commercial Internet infrastructures to

facilitate network-centric warfare (NCW) systems. On the positive side, this methodology

reduces costs and delivers a wide range of equipment quickly, thus increasing flexibility and

functionality. However, COTS purchases can lower reliability and limit opportunities to

verify that the software performs its stated purposes. It is more difficult for the buyer to

gain insight into source and application code documentation for COTS products especially if

the providing companies are offshore (Gansler and Binnenndijk 2004). Many COTS

components, and sometimes entire systems, are developed and maintained by providing

companies who may themselves procure development and services from other nations who

could have privacy, intellectual property rights, security, diplomatic, and defense policies at

odds with the original procuring country. Thus, a COTS strategy increases the possibility

that a hostile nation or non-government hostile agents (terrorist/criminal) could

compromise the system or services. Regardless of the trust level between the countries

and/or corporations, a single person working on military or critical infrastructure software

could cause havoc by installing programs that compromise combat zones, military and civil

command and control systems, and system access.

Offshoring significantly increases the risks to military systems because many network

components are produced and or shipped through countries that may be hostile to the

national interests of the procuring military organizations. Eugene Spafford, Chair of the

U.S. Public Policy Committee of the Association for Computing Machinery (USACM), testified

as follows in October 2005 before the House Armed Services Committee.

“Our military and government rely on COTS products and contractors to equip and

staff our IT infrastructure. Consider that some of those products that are employed

in highly sensitive applications are being crafted, tested, packaged and supported by

individuals who would never be allowed into the locations where those applications

are used because of national origin, criminal history, and/or personal behavior.

Furthermore, some of the hardware and software components in use in critical

applications are designed and produced in countries that may be adversaries in

future military or political conflict. These factors enable “life cycle” attacks where key

systems can be compromised during early manufacture, shipping, and maintenance

as well as end operation. We do not have the tools or resources to thoroughly check

these items to ensure that they do not have “hidden features” or flaws that may be

used against us. We need special attention and methods to produce these

supervisory systems and critical applications.” (Spafford 2005)

In Section 6.2, we discussed a recent Symantec threat assessment that shows rising

vulnerabilities and malware designed to compromise confidential information. Tying that to

the widespread use of COTS in most countries’ military systems, the national security risk

becomes clear. The effects of possible critical IT infrastructure breaches include, but are not

limited to, the following.

From a national security perspective, large-scale attacks that manipulate the

availability and integrity of military command and control systems can cause

malfunction, or in the worst case, loss of life, due to weapon trajectory changes

and battlefield misinformation.

Page 203

Disruption of IT-based systems and services can potentially increase a loss of

situational awareness of an attack, decreasing identification time and time to

respond. Not only could these attacks be catastrophic, recovery could be more

difficult if the deployed products were developed offshore and the capability to

manufacture and develop the hardware and software to replace non-trustworthy

or damaged systems is no longer available within the procuring country.

A significant example of a loss of national capability through global sourcing,

though not IT related, is the 2004 closing by UK health authorities of the sole US

supplier of a flu vaccine plant. This caused an immediate reaction in the United

States especially for citizens at risk. Agreements were obviously not in place

between the two countries to supply early warning of the vaccine contamination

(Stannard 2004). This experience is especially alarming in the context of growing

concern over the predicted bird flu pandemic. The lack of capability to

manufacture enough vaccine could lead to major political conflict between

procuring and providing countries and, in the worst case, massive fatalities could

occur as the virus spread, unabated by an antidote. This example illustrates how

technical capabilities can severely impact the ability of a nation-state to provide

for its populace.

Covert access to vital command-and-control systems could undermine military

strategies and battlefield success by either exposing or taking advantage of

military tactics or distorting data.

Unauthorized access to confidential records could leave military and civilian

personnel open to blackmail and other forms of compromise affecting national

security. Compromises causing intermittent failures or loss of integrity of data

can also affect loss of life on the battlefield. If failures were deliberately caused,

for instance, from built-in malware vulnerabilities such as trojans and bots, an

attack such as a buffer overflow or a web-based attack could allow the

undetected bypass of security mechanisms such as firewalls or virus scans.

Allowing access to remotely controlled bots for later attacks could undermine the

military and public/private infrastructure of these shared networks as well as

those in countries to which they connect.

A broad potential risk, one that could be considerably exacerbated by offshoring, is that

the providing organization (or at the least, significant parts of its ownership or

management) could be compromised and used by organized crime or foreign governments

to the detriment of organizations and citizens in the procuring country. This could serve as a

means for bringing about the negative consequences discussed previously. There are many

instances of businesses becoming beholden to organized crime interests or fronts for

government agencies in nations around the world. Globalization provides substantial new

opportunities and reach in this regard.

For such a scenario in an offshoring context, it is easy to imagine the providing

organization gaining control over data assets and management (e.g., databases, network

operations) that would give it a powerful platform to engage in such activities as

unauthorized data mining, intelligence operations, malware planting, attack planning, and

money laundering. Given its position as a provider, this might continue for a long period of

time, enabling it to do a great deal of damage in a relatively protected way.

In some countries, including Russia, parts of Eastern Europe, and China, among others,

where there has been a lack of a well-established rule of law that effectively protects

individuals or private enterprises, it may be difficult or impossible for provider organizations

Page 204

to resist overtures by organized crime or government security agencies. Potential providers

have very little means (physical or legal) to defend themselves against such overtures.

Critical Infrastructure – Operation and Investment

From a social perspective, attacks such as those described previously can also cause

malfunction and destruction of critical civilian infrastructure, for instance, transportation,

power, and financial systems, not to mention loss of civilian life, chaos, and loss of public

confidence in the national infrastructure and government. From an operational and

investment standpoint, it would be difficult to replace the aging backbone of domestic and

foreign-built equipment in the procuring countries’ infrastructure and also problematic to

train maintenance personnel to install, connect, and operate it. Thus, the impacts on

nuclear energy, electric, or water purification facilities could be detrimental not only to

health but also to the economy.

The UK Financial Services Authority (FSA) issued a report in May 2005 warning that

offshoring could damage consumer protection efforts and lead to increased financial crime

(Watson 2005). The report highlighted the greater difficulties of implementing strict controls

in offshoring. At the same time, the FSA stressed that risks could be addressed with

appropriate risk management strategies. The FSA noted that two key risks were business

continuity and high staff turnover.

The high turnover noted in the Financial Services Authority report is a security threat

because of the gap in personnel security noted in Section 6.2. Staff turnover has been a

problem in the rapidly expanding IT sectors of countries such as India because skilled staff

members are in high demand (NASSCOM-Evalueserve 2004). The high demand coupled

with the absence of searchable credit or criminal databases greatly increases the likelihood

of hiring higher risk employees.

Modern technology-based economies are highly dependent on an extensive array of ITcontrolled

infrastructure. IT-dependent systems include water and electric power,

emergency communications, transportation, oil and gas production and delivery, and health

care. These systems are vulnerable to hacking, sabotage, and natural disaster. Hurricane

Katrina and the New Orleans flood illustrates how quickly a situation can degenerate when

infrastructures fail. Supervisory Control and Data Acquisition Network (SCADA) Systems

control critical infrastructure facilities such as nuclear power plants, and these SCADA

systems are vulnerable to attack. Systems are not only vulnerable to attacks through their

non-Internet-based control systems, but through other, outdated control systems as well.

"They're designed to be managed remotely and the remote management is not

authenticated, meaning you don't know who's managing it,” according to Alan Paller, the

research director for the SANS Institute.” (Simmons 2005).

Offshoring introduces additional failure points into a system, and it also makes these

systems vulnerable to concerted attack in the event of hostilities.

Access and manipulation of financial and telecommunication systems could cause longterm

national and global economic damage which, if severe or frequent enough, could cause

loss of public confidence in infrastructures and the governments’ ability to protect the

population. More immediately, the economic costs of interrupted finance would likely

paralyze many modern societies. Examples are as simple as a local airline computer failure

that led to five hours of queuing to board one international flight (British Air Flight 286, San

Francisco to London, March 5, 2005) in order to achieve correlation of baggage with

passengers. The September 11 terrorist attacks came within an estimated one mile of

destroying the redundant backup for the Eastern Seaboard Point of Presence, the hub

concentrator for both voice and web-based traffic for 100 million people, and nearly 65% of

America’s financial transaction backbone. On the other hand, the Internet protocol proved

its survivability under the duress of these events. Although NYSERNet’s research network

Page 205

ran through Ground Zero with transport provided by Verizon, on lines that were severed

when Building 7 of the World Trade Center collapsed, the network never wavered.

Moreover, the technology’s flexibility helped restore commodity service on Long Island and

in Westchester County by remapping onto the NYSERNet network (Lance 2002).

The September 11 terrorist attack was a warning flag to the US financial infrastructure.

The damage, as incredible as it was, could have been much worse. US regulatory agencies

took steps to increase scrutiny of offshoring of the financial services electronic commerce

systems. The US financial system has become a worldwide network, and along with the

Security and Exchange Commission, the US Federal Financial Institutions Examiners Council

(FFIEC) began requiring audits of offshore providers. FFIEC issued regulations designed to

increase the scrutiny of offshore service providers.

National Economic Health, Security, and Outsourcing

For most nations, creating and maintaining a robust economy with adequate jobs is a key

to a sound and stable government. Governments are supported because citizens believe

their leaders’ policies best serve their national defense, economic, and social interests.

When confidence in government with regard to these factors falters significantly, instability

is usually not be far behind. A country’s national security and social policies are influenced

by its technological development, natural resource availability and utilization, strength of

defense, soundness of financial and operational infrastructure such as transportation and

energy systems, trade policies, citizens’ ability to create and innovate, and cultural and

historic heritage. However, this discussion will focus primarily on issues that may be

affected by offshoring.

Offshoring exacerbates some old issues and raises new ones. The rapidly changing

technology transfer associated with offshoring has already begun to change country

objectives, policies, and cultures for both procuring and providing countries. In recent

historical experience, the procuring nation has rarely feared the loss of economic and

intellectual advantage to the providing country, yet, as manufacturing industries have

shifted jobs away from North America and Europe, that is precisely the concern for those

concerned about offshoring.

A key concern for the United States and other developed countries is whether their

technological investment and innovation will decline so steeply as to put them into economic

decline. The question therefore is how the developed countries that have been technology

leaders can preserve their technological capabilities for innovation to maintain successful

economies even if they are no longer broadly preeminent over their rivals. There are

numerous studies, reports, and commentaries making these points (see, for example,

National Summit on Competitiveness 2005; Harsha 2005; National Academies 2005; Lewis


One example is the testimony given by Nicholas Donofrio, Executive VP for Innovation and

Technology at IBM, before the House Science Committee. Donofrio called for a national

innovation ecosystem that he said must be fostered by a coherent national policy. In his

testimony, he quoted from the December 2004 Report of the National Innovation Initiative

and it is worthwhile to review it in some detail. “The push and pull of supply and demand

do not occur in a vacuum. They are strongly influenced by public policy and the overall

infrastructure for innovation offered by our society. Public policies related to education and

training, research funding, regulation, fiscal and monetary tools, intellectual property and

market access demonstrably affect our ability to generate innovation inputs and respond to

innovation demands. The same can be said of infrastructure – be it transportation, energy,

health care, information technology networks or communications. Taken together, the policy

and infrastructure environments create a national platform that can accelerate – or impede

– the pace and quality of innovation.” (Donofrio 2005)

Page 206

At What Point Does Declining IT Capability Impact National Security?

The United States, which has offshored significant amounts of manufacturing capability

over the past half century, is increasingly offshoring its IT capability. Does this really

matter? Consumer electronics went offshore without much ill effect on either the national

economy or the national security in the 1980s. So did memory chips for computers, and

after that, the PC’s themselves, plus their displays. Some economists argue that, in fact,

this offshoring led to large productivity gains for the United States (Mann 2003, 2004,


Many argue that software is different; it does not act as a commodity since the

development cycle never ceases. There is often an innovative edge tied to market

differentiation that moves well beyond the current frontier of commodity service. This is the

uniquely special characteristic of software compared to any other building material. If a

country loses control of that frontier, does it risk losing control of its future in both national

security systems and in many critical sectors such as financial services, health care, utilities,

and industrial controls? Since the critical capability of market differentiation and the agility

of systems depend on software capability, the answer is neither easy nor encouraging for

US planners.

The future economic welfare of the United States (or other developed countries) could be

at risk if offshoring, combined with the absence of appropriate policies, damage the nation’s

ability to produce technological innovation. It is competitive products and higher

productivity that lead most directly to national wealth, but many observers believe that

technical innovation is the underpinning for competitive products and higher productivity in

the knowledge economy. The concern is that the decline in investment in American

research and development and education spending could accelerate to the point where it

jeopardizes future reproduction of intellectual capital. This topic is discussed in detail in

Chapter 8.

Several popular books explore this issue of national risks from offshoring, one by

journalist Thomas Friedman, another by a former trade official in President Reagan’s

administration, Clyde Prestowitz. These books present visions of the future, though with

less than academic rigor. In The World is Flat, Friedman explores the results of a highly

interconnected globalized world, driven hard by offshoring of IT and its effects on the United

States (Friedman 2005). The result, in Friedman’s view, is an inexorable and extremely

rapid shift of many presumed US advantages, most especially white-collar jobs, to other

countries. He predicts a shock wave impact on American politics, business success, and

unemployment rates that will likely result in clarion calls for revised educational approaches

and tariff legislation as well as much acrimony. But he stops short of saying either that it is

a momentum that can be stopped, or for that matter, needs to be

In Three Billion New Capitalists: The Great Shift of Wealth and Power to the East,

Prestowitz argues that the United States faces such serious fiscal and competitive

challenges that it may be headed not only for a declining standard of living but for a 1930'sstyle

depression. The subtitle for Prestowitz’s book is telling in itself: The End of Western

Dominance and the Rise of Parity (Prestowitz 2005). Prestowitz is gravely concerned

because the United States is not prepared for the expected economic restructuring driven

by globalization. He refers to the mismanagement of the US economy, manifest in low

household savings, high budgets shortfalls, and unsustainable trade deficits and foreign

borrowing. A deeper problem for Prestowitz is the fact that the United States has no

national strategy to protect its industry, skilled workers, and technological leadership. He

argues that the United States’s laissez faire economic ideology and confidence in its

technological and productivity supremacy have prevented Washington from grasping the

coming crisis and from developing a programmatic national response.

Page 207

Dealing with Risks and Exposures in National Capability

The risks and exposures from IT offshoring are great and increasing. They include

increasing cyber vulnerabilities, hostile cyber warfare policies, theft and abuse of personal

and government sensitive and classified information in all countries, attacks on country

infrastructure, and changes in business strategies and investment in research and

development. Globalization is likely to continue and so are its international effects. These

risks and exposures can never be completely mitigated, but strategies at both the national

and international levels can be put in place to help manage them.

Problems cannot be solved until they are defined and accepted as valid by a sovereign

entity and its citizens. Frank and open national dialogue regarding economics, trade,

outsourcing, education, and research issues that does not focus on a corporate or protective

agenda would allow citizens to engage in the dialogue and understand the issues.

One topic that the United States and other developed countries might address are plans

to protect their nation’s cyber-structure and IT competitiveness. The plan might include not

only a strategy to address training and jobs but also strategies for legislation, international

agreements, policing, tariffs, Internet policies, and a more equitable tax-structure for

companies investing at home. It might address the need for more formal

government/commercial agreements and funded research to address data protection and

communications between stakeholders involved in homeland defense and critical

infrastructure. It might also include a discussion of how to make a country more

innovative, specifically in light of offshoring which is discussed in detail in Chapter 8.

The offshoring of homeland security technology development and management systems

that send vital information such as biometrics, identification codes, tax and personal

information overseas are of critical concern. Until better controls for this information are

developed, it presents a high risk to all nations. Sensitive industries should have severe

restrictions on offshoring. Offshoring of software and design projects in areas such as

defense and the other critical infrastructure industries should be tightly controlled. Further

research in methods to secure this data and the development of nation-to-nation and

international treatment of both the data and how compromises will be handled is vital,

including developing and implementing information security standards for international


Thomas Homer Dixon, the Director of the Trudeau Peace and Conflict Studies Center at

the University of Toronto, has studied the relationship between violent conflicts and various

kinds of environmental stress in poor countries. He found that environmental stress cannot,

by itself, cause violence. It must combine with other factors, usually the failure of economic

institutions and government. He concluded that a central feature of societies that adapt

well is their ability to produce and deliver useful ideas or what he calls ingenuity to meet the

demands placed on them by a worsening environment. Societies that adapt well are those

able to deliver the right kind of ingenuity, at the right time and places, to prevent

environmental problems from causing severe hardship and, ultimately, violence. If

globalization is to be successful, recognition of the rights and equality of the global citizen

has to be accepted and become the underpinning of policies and trade agreements. Dixon

speculates on whether procuring countries and providing countries can learn from history

and forgive past injustices and whether this might determine if global and worldwide

innovation will continue.

6.6. Risk Mitigation and Risk Assessment

A basic approach to information security risk assessment is to analyze three key

objectives: confidentiality, integrity, and availability. The risks to these objectives are

Page 208

greatly increased by offshoring because of inherent vulnerabilities in offshoring, global

communications, and international business.

To illustrate, consider the example of ABG, a fictional software company that sells

equipment for processing secure transactions. ABG’s systems are only valuable to

customers if (1) the internal security mechanisms are kept confidential from competitors

and potential attackers; (2) the integrity of transactions are ensured (the data cannot be

changed by an attacker); and (3) the process is available and efficient, that is, the process

does not slow down or interfere with the client’s primary business. Outsourcing the

development of new features for the product increases the risks of a competitor or a

potential attacker learning ABG’s proprietary processes. It also increases the risk that the

process will not be as reliable (or have as much integrity) due to the loss of control over the

development process and a more complicated supply chain network. While any outsourcing

increases these risks, developing the software in another country magnifies them

significantly for a number of reasons. ABG no longer controls the network security or the

process security of the development center. Offshore developers likely have less legal

liability to ABG or its clients. The development and maintenance processes (discussion of

proprietary designs, transfer of the software, patches, documentation) are conducted over

international global networks with a greater potential for interception. A prime motivator

for both provider and procurer becomes cost reduction, which tends to overshadow security

or other quality concerns

Risks exist at multiple levels – financial, performance, reputation, intellectual property,

privacy, legal, and regulatory risks. Therefore, it is imperative to carefully assess the risks,

quantify the potential losses, and develop cost-effective risk mitigation strategies.

Systematic risk analysis and planning involves the following steps.

Identify and estimate the value of the assets to be protected.

Identify the potential threats (things that can go wrong) and threat perpetrators.

Assess the vulnerabilities in the current systems protecting the assets.

Develop a plan to protect the assets against the threats by remediation of


Too often, providing companies simply hire guards and put expensive firewalls and access

controls on their network, and then they declare to the procurers that there is no risk. The

managers of the offshoring programs at procuring companies repeat this mantra. However,

if the asset being protected is client privacy or corporate intellectual property, all it takes is

a disgruntled or dishonest employee to copy the data and walk out of the well-protected

offshore data center and sell the information to the highest bidder.

Recognition of higher risks for the procurer is not necessarily an argument against

offshoring. Risks lead to innovation, and the free market is based on the principle of taking

and overcoming higher risks to obtain higher rewards. The accepted response to a risk is

avoidance, transference, or mitigation, responses all found within industry for managing the

risk profiles of offshoring. The key is a conscious rational assessment and response to the

risks in each situation.

Mitigation Strategies

Effective risk mitigation strategies need to be implemented once a risk assessment exists.

Security due diligence. This should include certification to standards such as

BS7799, CISP, or True Secure. (Note that SAS70 is not effective in info security

space. Note also that Certification cannot be viewed as a license to ignore

Page 209

security risks.) Legal liability and responsibility for protecting both customer data

and intellectual property lie with the procuring company management.

Business due diligence. Does the provider have the technical and security skills

needed? Does it conduct effective background checks? Is it financially stable?

What relationships does it have with other companies, governments, and


Active risk management. This requires the development and implementation of

an ongoing security plan between the outsourcing procuring company and the

provider. The plan should include appropriate forms of monitoring, regular

reporting of security metrics, incident response, and disaster recovery


Third-party risk assessments. An independent third party should be responsible

for regular security audits of the provider. This should be a professional security

firm rather than an auditing firm. Past practice has used the CPA issued SAS70

audit as an acceptable security assessment and audit. However, SAS70 audits

are usually produced by the audit firm employed by the provider. The scope of

the audit is generally defined by the provider and typically includes a pre-audit

that allows the provider to correct any embarrassing failures. Consequently, real

risks are often overlooked. Therefore, it is essential that a procuring company

ask for an independent security assessment.

6.7 Bibliography

Alexandrov, V.N. 2005. (Ex-Soviet CS Director, today Chief o