21

copy 2004 The Massachusetts Institute of TechnologyInformation Technologies and International Development

Volume 1 Number 2 Winter 2003 21ndash 37

Raregq DossanidossanistanfordeduAsiaPacic Research CenterStanford UniversityE309 Encina Hall616 Serra StreetStanford CA 94305-6055

Martin KenneymfkenneyucdaviseduDepartment of Human andCommunity DevelopmentUniversity of California Davis1 Shields AveDavis CA 95616

Berkeley Roundtable on theInternational EconomyUniversity of CaliforniaBerkeleyBerkeley CA 94720

Lift and Sh ift Dossan i Kenney

ldquoLift and Shiftrdquo Moving the BackOffice to India

The recent growth in offshoring business processes is driven by the need for

cost savings but because of the potential for both the quantity and quality of

work that may be done overseas has larger implications for the service

economy in developed countries This paper uses India as a case study to

examine the business knowledge-related and technological considerations

that drive the globalization of business process fulllment It also examines the

industrial structure that is emerging in India for the work and draws

conclusions about its future and its implications for service workers in

developed countries

Today the physical products we consume are the result of a value chain inwhich much of the manufacturing is undertaken offshore (Feenstra 1998Geref 1994 Porter 1990) During the last three decades the operatingassumption has been that this loss of manufacturing is natural as we in-creasingly become a service economy Though usually not explicitly articu-lated sanguine observers implicitly assumed that services would remain inthe developed nations with their more highly educated workforces supe-rior telecommunications and transportation infrastructures and moretransparent government and regulatory regimes1 This with only minor ex-ceptions characterized the global geography of work before the early1990s

The world of services is most complicated because it contains such awide variety of activities Many services such as hairstyling coaching andhouse painting to name only a few of what Reich (1991) terms in-personservices are place based There are several other services such as data en-try accounting and claims processing to provide some illustrations thatare not as place based For these less rooted services there may be verygood reasons they are located close to their actual generation but underthe correct conditions these reasons might become less binding For ex-ample transportation or communications barriers may exist in terms of ei-ther cost or security The information necessary to carry out the servicemay be embedded in physical materials that cannot be efciently or se-curely relocated That is not to say that no services have been offshored

The authors thank Frank Mayadas Gail Pesyna and the Alfred P Sloan Foundation for funding the research that con-tributed to the preparation of this paper We also thank the referees for their comments as well as NASSCOM Indiaand the Software Technology Parks of India for their assistance The authors remain solely responsible for any conclu-sions and opinions expressed in the paper1 For an early discussion of the movement of the United States to a service economy see Bell (1973) For concernabout the loss of manufacturing see Cohen and Zysman (1987)

Already in the 1980s some English-speaking Carib-bean nations were undertaking airline ticket andcredit card processing (Pearson 1993 Posthuma1987) for various multinational corporations (MNCs)In the 1980s airlines began ying ticket stubs to Ire-land or India for processing (Warf 1995) Still thenumber of overseas employees in these operationshas been small and for an economy the size theUnited States the effect has been minimal Not sur-prising given the massive movement of manufactur-ing jobs offshore these small operations haveattracted little attention

In the 1990s the pace of service offshoring ac-celerated One major beneciary of this was Irelandwhich became a haven for US software MNCs thatwere able to take advantage of Irish tax breaks andits well-trained inexpensive workforce to providecall center services and software indigenization forthe European market (OrsquoRiain 2004) Contempora-neously in India a contract software programmingindustry servicing MNCs especially in the UnitedStates and the United Kingdom emerged (A Aroraand Athreye 2002 DrsquoCosta 2003) These develop-ments indicated that services could also be under-taken offshore In the last half of the 1990s a fewMNCs began experimenting with relocating businessprocesses (BPs) that is components of back-ofceservices such as payroll and order fulllment andsome front-ofce services such as customer care toEnglish-speaking developing nations especially In-dia but also other nations such as the Philippines

Besides India many nations including Malaysiathe Philippines China the English-speaking Carib-bean and Ireland are courting MNCs with the hopeof attracting these BP operations However the In-dian BP operations already employ more than all theothers combined Each of these other nations hasdisadvantages compared with India For example inMalaysia the Philippines and the Caribbean the la-bor forces are too small In China English languageremains a difculty Only the Philippines appears tobe a strong competitor and it has a reputation forbeing politically unstable For this reason this paperdiscusses only India

In recent months the popular press has been fullof stories about how white-collar jobs in the UnitedStates are at risk of being transferred overseas Forexample Engardio Bernstein and Kripalani (2003)had a front-page cover in Business Week asking thequestion ldquoIs Your Job Nextrdquo which intimated that

the entire white-collar class was at risk The othernews media have chimed in with similar articleseven as liberals have denounced this new ldquohollow-ing outrdquo of the US economy There is naturallymuch concern about the implications of moving BPsto India There is however little substantive infor-mation on the industry with the exception of con-sultant reports and newspaper articles This articleaims to provide some baseline facts about the indus-try and provide a descriptive analysis of the currentindustrial structure of the Indian BP industry

The paper begins with a discussion of our re-search methodology and how we collected dataBecause there is little written on this industry exceptnewspaper articles and consultantsrsquo reports weused personal interviews as the main data-gatheringtechnique The next section explains what BPoffshoring is and examines the technical and busi-ness enabling conditions that made the phenome-non not only possible but also attractive to MNCsWe also identify the six factors that rms exploringoffshoring BPs consider before making such a deci-sion The third section discusses the drivers of thedecision to locate BPs in India In the fourth sectionthe industrial structure of the BP industry in India isexamined and seven organizational forms are foundto be coexisting at the present time Then in thediscussion section we draw some tentative conclu-sions about the future of the industry in India andits implications for service workers in the developedcountries

This study reports on exploratory research that theauthors conducted at business process operations inMumbai New Delhi and Bangalore in March andApril 2003 The sample of 46 rms comes from atotal population of approximately 200 BP rms in In-dia We asked the two Indian organizations most re-sponsible for the IT sector the Science andTechnology Parks of India and NASSCOM to choosea representative sample for us However we deliber-ately oversampled on the larger more signicantrms We conducted interviews with 46 rms thattogether account for 40 of the total offshore BPemployment in India (see Table 1) We interviewedrepresentative rms in each of the six rm catego-ries identied in our research Though we concen-trated on the largest rms we also visited

22 Information Technologies and International Development

Lift and Shift

operations as small as 100 employees Hence webelieve that the rms we interviewed are representa-tive of the Indian BP offshoring industry

The interviewees were executives usually vicepresidents with various functional titles and infewer cases the CEOs or chairpersons The inter-views were invariably at least 1 hour in length andoften up to 2 hours In 70 of the cases there wasa tour of the facilities though because of the timedifference with the United States the facilities usu-ally had minimal activity All interviews were tapedand transcribed Unfortunately because of the polit-ical and commercial sensitivity of business processoffshoring the respondents requested anonymity

In addition to the interviews we reviewed the lit-erature on BP offshoring and outsourcing This liter-ature falls in two realms consultant reports and thepopular press The consultant reports can be valu-able but frequently suffer from their desire to createbusiness for themselves The press reports are usefulfor providing topical materials especially in follow-ing investments and expansions However exceptfor these two sources of written material there is lit-tle other material on the topic

Business processes is the catchall term used for themyriad white-collar processes that any bureaucraticentity undertakes in servicing its employees ven-dors and customers2 These include human re-sources accounting auditing customer caretelemarketing tax preparation claims processingdocument management and many other chores

necessary for rm functioning Invariably the com-pletion of a BP is the result of an entire chain ofseparate activities

Until the early 1990s business processes weregenerally treated as a xed cost and received mini-mal management attention The interest inreengineering awakened management attention tothe savings that could be achieved by reorganizingthese processes (Hammer 1990 Hammer andChampy 1993 Davenport 1993 Cole 1994) Onepart counseled the decomposition examination andstandardization of the activities necessary to com-plete a process This was often accompanied by adigitization of at least some portions of the entireprocess The reengineering exercise made each activ-ity in a process more visible and often certain activi-ties were modularized thereby permitting detailedconsideration of the most cost-effective way ofcompleting each activity and making it possible toseparate them spatially and organizationally shouldthe need arise3

This reengineering was combined with the in-creasing digitization of the objects of service workthat was ongoing throughout the 1990s Digiti-zation was so important because it removed thephysical medium carrying the information thus free-ing it from the constraints of size and weight(Kenney 1997 Curry 1997) Still few service opera-tions can be done only on the computer (the mod-ern form of mundane work in the terminology ofCallaghan Thompson and Warhurst [2001]) Theyusually require some level of face-to-faceinteractivity either among coworkers or with per-sons outside the organizations such as vendors andclients However information technology (IT) and

Volume 1 Number 2 Winter 2003 23

Dossani Kenney

2 We dene a business process as a complete service such as handling a customer complaint processing a medicalclaim or processing a purchase order Completing a process requires undertaking a set of activities For example inhandling a customer complaint it is necessary to understand the complaint classify it decide on a course of action un-dertake the action and follow up to ensure the action solved the complaint Each of these is an activity that is poten-tially separable from the others3 On modularity see for example Baldwin and Clark (2000) and Sturgeon (2002)

Table 1 Sample Firm Size

No of rmsinterviewed

Employmentin sampled

rms

Totalindustry

employment

Sample ofindustry

employment

Averageemployees

per rm

Medianemployees

per rm

42 69729 171500 407 1660 1000

reengineering permitted processes to be parsed intocomponents requiring different levels of skill andinteractivity This made it more feasible to relocatethose activities requiring low levels of face-to-faceinteractivity offshore Orlikowski (1996) captures thisphenomenon in her study of technical help deskswhere digitization induces work ows to be reorga-nized in the direction of dividing work ow intotasks that are separable in terms of technical skillsand interactivity These developments had signicantimplications for the possibility of offshoring even inthe absence of cost arbitrage opportunities that areavailable when work is outsourced to developingnations such as India

The initial relocation of BPs was intranational Forexample call centers were relocated from high-costcities to lower-cost small towns and then to rural lo-cations The earlier efforts to lower wage costs forback-ofce BPs saw rms move their back-ofce op-erations to smaller Midwestern towns where accentswere neutral education was adequate and laborcosts were lower and (at that time) relatively reliableThe cost savings were likely in the 20 to 30range when the low-cost land and labor costs weretotaled However labor pools were shallow thuslimiting the scale of operations

From this perspective the relocation of suchwork to other countries was an extension of thisprocess as rms sought to cut costs further Therewere technical developments that facilitated reloca-tion One of the most important was the continuouscost reduction in data transmission These technicaldevelopments loosened the geographical constraintsthat encouraged the undertaking of BPs nationallyReducing the physical and technical constraintsmeant that variables such as factor costs and skillavailability increasingly became the determinants ofwhere service activities were undertaken

In choosing which processes to undertake off-shore it may be thought that the simplest processeswould tend to be offshored rst because the skillsfor undertaking more complex processes might takea longer time to learn Most existing forms of glob-alization are not processual and thus do not consid-er the evolution of the overseas facility however asKogut and Zander (1992) among others hypothe-size overseas operations also experience learningeven if it is simply catch-up From the perspective ofthe source nation the initial transfer may be onlythe rst stage in the absorption of an entire chain of

activities For example in insurance claimsprocessing whereas initially the Indian operationmight simply process the claim it might be possibleto transfer some of the investigation and valuationactivities to the Indian operation With experienceIndian accountants or engineers may be trained todetermine fraud or exaggeration of claims or at theleast ag unusual claims

There are several additional dimensions to theoffshoring decisions other than the complexity oftasks and the need for face-to-face interactionSome of these are internal to the rm and some arepart of the environment It may happenmdash as wehave noticed in Indiamdashthat highly technical com-plex tasks in customer support that require interac-tion between say a corporate client in the UnitedStates and the vendorrsquos technician may be moreeasily offshored than a simpler task requiring inter-action with a retail customer This is because the lat-ter requires a kind of interactive skill that might beweaker in India compared with the former whereasIndia offers technical skills of high quality capable ofhandling the former task

We identify the following six internal factors thatrms consider before offshoring an activity in addi-tion to environmental factors such as technologicalenablers and the savings in direct labor costs (whichwe consider later)

The knowledge component of the activity Ahigher knowledge component makes the rmmore concerned about whether the quality ofthe service will change because of a locationalchange and greater difculty in the transferprocess There may be reasons to worry aboutquality slippage if the remote location cannotunderstand the quality or qualities needed or(even if it understands) cannot match the qual-ity needed Sometimes if quality is acceptablea rm may not only shift to a remote locationbut may also outsource the activity to mitigatethe possible political difculties of an internaltransfer A rm also might be concerned abouta loss of competencies in a certain locationthat would be costly (or even impossible) to re-acquire in the same location if ever needed Ifunique skills atrophy any disruption of accesscould have dire consequences

The interactive components of the processThere are two dimensions of interactivity that

24 Information Technologies and International Development

Lift and Shift

affect the location of BPs (1) interaction withother persons in the production of the serviceand (2) interaction with the (clientrsquos) customerand other vendors The greater the need to in-teract with other persons within the rm andthe greater the interaction across different ser-vices and processes the higher the cost andrisk threshold of offshoring portions of thework Of course it might be possible to off-shore the entire process thus retaininginteractivity at the new location But if someactivities cannot be offshored offshoring theothers might not be feasible In cases whereoffshoring and outsourcing were both beingconsidered the risks might be even greater Forcertain rms customer interaction is a coreskill If they believe offshoring could disruptsuch interaction they are unlikely to take therisk

The level of separability of the process In cer-tain kinds of work such as customer care theentire process can be offshored because it doesnot require personal interaction These activitiesusually are routinized and even scripted For ex-ample much software support is providedthrough scripted decision trees For lessscripted activities offshoring may be less feasi-ble Consider a process within which 70 ofthe work is Web or database research and30 is interacting directly with clients Unlessit is possible to separate these processes intoactivities it will be impossible to transfer theprocess Separability is not dichotomous butrather a matter of degrees It may be possibleto reengineer a process to make it more sepa-rable though this may be at the cost of somelevel of service

Savings from concentrating an activity in onelocation Often rms will have several ofcesperforming the same function This is often thecase because as rms expand they often tapout local labor pools and therefore establishofces in other regions This can result in aninefcient spatial posture Often in developedcountries it is too expensive to concentratethese facilities in a single location Howeverthe rm relocating to India can overcome thissunk-cost issue The advantages of concentra-tion may be signicant

There are benets of size due to scaling upoperations For example a larger facility ismore capable of operating overload A largerfacility requires a smaller excess cushion tomanage a peak load thereby creating both amanpower and facility savings Of course theremay also be diseconomies of size The advan-tage of concentration can also favor out-sourcing because a service provider can poolthe business of many clients The classic exam-ple is medical transcription which is still mainlyoutsourced to small local rms or even individ-uals in the United States The large rms in In-dia can offer guarantees of quality that smallerdomestic operations cannot offer because theyhave large labor pools thus reducing the ef-fects of absenteeism and offering the typicalefciencies that Adam Smith ([1776] 2003)points out regarding even a simple division oflabor

Reengineering as part of the transfer processTo transfer a business process it is necessary tostudy it intensively and script the transfer Inthe process of study often there will be stepsof the current process that can be discardedconsolidated or simplied These are legaciesof earlier methodologies that were changed asthe production process evolved During thetransfer these are easier to abandon than at anexisting facility where they have become a nat-ural component of the daily routine There areother less expected benets that can arise fromthe transfer process that go beyond theefciency effects These actions of examiningand transferring a process can yield signicantefciencies Processes often evolve in a path-dependent manner and as Arthur (1994) andDavid (1986) show path-dependent outcomesmay not be the most efcient conguration ve-hicle During the transfer process theseinefciencies can be addressed without disrupt-ing work patterns The workers in the new lo-cation are met with a fait accompli

The time-sensitive nature of the work Somework such as payroll has deadlines Indian op-erations can shorten various product cyclesThe most obvious of these is work that needsto be completed overnight in Europe or theUnited States that can be undertaken during

Volume 1 Number 2 Winter 2003 25

Dossani Kenney

the day in India For example technical calls byhigh-end clients may need to be answeredwithin 15 minutes For medical transcriptiondoctors notes for patients in intensive caremight have turnarounds as short as 2 hoursIndian operations are able to undertake thesehigh-priority tasks because they can affordmore slack resources to meet peak loads thantheir Western counterparts can This dimensionof quality is far more expensive to offer in de-veloped nations where labor is more costly

The combination of low labor costs project man-agement skills and technological sophisticationmake India a particularly attractive candidate forBP offshoring4 The genesis of the relocation ofbusiness processes to India can be traced to themid-1980s when India emerged as an offshoresite for software production By the late 1990s In-dia had become a leading supplier of contract soft-ware programming because of its combination ofskilled low-cost manpower and project manage-ment skills

Labor cost differentials are not the only differ-ence In the United States call centers have beenrelocated to low-labor-cost areas which are mostoften rural areas There they are limited by laborshortages high attrition rates and relatively pooreducational levels In contrast Indian cities such asMumbai or Bangalore offer access to enormouslabor pools of potential employees and have thecomplementary urban services that permit the estab-lishment of operations employing thousands Thecall centers in India we visited varied in size but themedian was 1000 employees (see Table 1) The av-erage size in the United States is under 400 andmany are between 150 and 300 employees

The key benets for foreign rms to operate inIndia are cost and labor supply According to an In-dian trade group NASSCOM-McKinsey (2003) BPoffshoring employment for overseas customers was171500 as of March 2003 compared with 106500in March 20025 Revenue generated for the nancial

year ending March 2003 was estimated at $2375billion compared with $1475 billion in the previousyear This implies that revenue per employee was$13848 in 2002ndash 2003 unchanged from the pre-vious year

According to the NASSCOM estimates directcosts per employee are $10354 (or $520 per bill-able hour of which $310 is the estimated laborcost) yielding a 337 return on the direct cost ofeach employee seat Comparably direct costs in theUnited States are estimated by NASSCOM at$55598 per employee (or $2780 per billable hourof which $2150 is the estimated labor cost)6

Clearly the cost pressure to shift the business over-seas especially during the current economic down-turn must be great

Because the median size of rm is 1000 persons(see Table 1) with typical xed-cost investment of$7500 per seat the annual return on capital is466 and average gross prot (which is tax ex-empt under current law) is about $35 million OneFortune 500 rm that consolidated severalfulllment operations to Bangalore reported to usthat the overall cost savings were 807 These rep-resent signicant dollar savings The NASSCOM-McKinsey report (2003) nds that General Electricone of the pioneers of outsourcing service opera-tions to India had achieved an annual savings of$340 million per year from its Indian operationsnow 7 years old Even if these numbers are inatedthe savings are remarkable and accrue directly to thermrsquos protability8

The NASSCOM-McKinsey (2003) report estimatesare that the Information Technology Enabled Ser-vices (ITES) operations will employ between 900000and 1000000 people in 2008 This assumes a com-pound rate of growth of approximately 45 peryear for the next 5 years an assumption yet to bevalidated This is possible because India has thelabor pool the enabling technologies exist and areimproving and the pressure on rms in developednations to lower costs is great If the current trajec-tory continues offshoring may prove to be of greatsignicance to nations and regions that could expe-

26 Information Technologies and International Development

Lift and Shift

4 For an early discussion of the role of India in BP offshoring see Aron and Singh (2002)5 Data for this section draw on NASSCOM-McKinsey (2003) pp 56ndash716 Note that this shows that infrastructure costs are lower in India This was probably always true for most infrastruc-ture though the plummeting cost of data communications (PCs software and telecommunications) to global levels isa recent outcome of reform as is the low cost of nance7 Personal interview by authors April 20038 The authorsrsquo interviews appear to validate the savings potential

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

Already in the 1980s some English-speaking Carib-bean nations were undertaking airline ticket andcredit card processing (Pearson 1993 Posthuma1987) for various multinational corporations (MNCs)In the 1980s airlines began ying ticket stubs to Ire-land or India for processing (Warf 1995) Still thenumber of overseas employees in these operationshas been small and for an economy the size theUnited States the effect has been minimal Not sur-prising given the massive movement of manufactur-ing jobs offshore these small operations haveattracted little attention

In the 1990s the pace of service offshoring ac-celerated One major beneciary of this was Irelandwhich became a haven for US software MNCs thatwere able to take advantage of Irish tax breaks andits well-trained inexpensive workforce to providecall center services and software indigenization forthe European market (OrsquoRiain 2004) Contempora-neously in India a contract software programmingindustry servicing MNCs especially in the UnitedStates and the United Kingdom emerged (A Aroraand Athreye 2002 DrsquoCosta 2003) These develop-ments indicated that services could also be under-taken offshore In the last half of the 1990s a fewMNCs began experimenting with relocating businessprocesses (BPs) that is components of back-ofceservices such as payroll and order fulllment andsome front-ofce services such as customer care toEnglish-speaking developing nations especially In-dia but also other nations such as the Philippines

Besides India many nations including Malaysiathe Philippines China the English-speaking Carib-bean and Ireland are courting MNCs with the hopeof attracting these BP operations However the In-dian BP operations already employ more than all theothers combined Each of these other nations hasdisadvantages compared with India For example inMalaysia the Philippines and the Caribbean the la-bor forces are too small In China English languageremains a difculty Only the Philippines appears tobe a strong competitor and it has a reputation forbeing politically unstable For this reason this paperdiscusses only India

In recent months the popular press has been fullof stories about how white-collar jobs in the UnitedStates are at risk of being transferred overseas Forexample Engardio Bernstein and Kripalani (2003)had a front-page cover in Business Week asking thequestion ldquoIs Your Job Nextrdquo which intimated that

the entire white-collar class was at risk The othernews media have chimed in with similar articleseven as liberals have denounced this new ldquohollow-ing outrdquo of the US economy There is naturallymuch concern about the implications of moving BPsto India There is however little substantive infor-mation on the industry with the exception of con-sultant reports and newspaper articles This articleaims to provide some baseline facts about the indus-try and provide a descriptive analysis of the currentindustrial structure of the Indian BP industry

The paper begins with a discussion of our re-search methodology and how we collected dataBecause there is little written on this industry exceptnewspaper articles and consultantsrsquo reports weused personal interviews as the main data-gatheringtechnique The next section explains what BPoffshoring is and examines the technical and busi-ness enabling conditions that made the phenome-non not only possible but also attractive to MNCsWe also identify the six factors that rms exploringoffshoring BPs consider before making such a deci-sion The third section discusses the drivers of thedecision to locate BPs in India In the fourth sectionthe industrial structure of the BP industry in India isexamined and seven organizational forms are foundto be coexisting at the present time Then in thediscussion section we draw some tentative conclu-sions about the future of the industry in India andits implications for service workers in the developedcountries

This study reports on exploratory research that theauthors conducted at business process operations inMumbai New Delhi and Bangalore in March andApril 2003 The sample of 46 rms comes from atotal population of approximately 200 BP rms in In-dia We asked the two Indian organizations most re-sponsible for the IT sector the Science andTechnology Parks of India and NASSCOM to choosea representative sample for us However we deliber-ately oversampled on the larger more signicantrms We conducted interviews with 46 rms thattogether account for 40 of the total offshore BPemployment in India (see Table 1) We interviewedrepresentative rms in each of the six rm catego-ries identied in our research Though we concen-trated on the largest rms we also visited

22 Information Technologies and International Development

Lift and Shift

operations as small as 100 employees Hence webelieve that the rms we interviewed are representa-tive of the Indian BP offshoring industry

The interviewees were executives usually vicepresidents with various functional titles and infewer cases the CEOs or chairpersons The inter-views were invariably at least 1 hour in length andoften up to 2 hours In 70 of the cases there wasa tour of the facilities though because of the timedifference with the United States the facilities usu-ally had minimal activity All interviews were tapedand transcribed Unfortunately because of the polit-ical and commercial sensitivity of business processoffshoring the respondents requested anonymity

In addition to the interviews we reviewed the lit-erature on BP offshoring and outsourcing This liter-ature falls in two realms consultant reports and thepopular press The consultant reports can be valu-able but frequently suffer from their desire to createbusiness for themselves The press reports are usefulfor providing topical materials especially in follow-ing investments and expansions However exceptfor these two sources of written material there is lit-tle other material on the topic

Business processes is the catchall term used for themyriad white-collar processes that any bureaucraticentity undertakes in servicing its employees ven-dors and customers2 These include human re-sources accounting auditing customer caretelemarketing tax preparation claims processingdocument management and many other chores

necessary for rm functioning Invariably the com-pletion of a BP is the result of an entire chain ofseparate activities

Until the early 1990s business processes weregenerally treated as a xed cost and received mini-mal management attention The interest inreengineering awakened management attention tothe savings that could be achieved by reorganizingthese processes (Hammer 1990 Hammer andChampy 1993 Davenport 1993 Cole 1994) Onepart counseled the decomposition examination andstandardization of the activities necessary to com-plete a process This was often accompanied by adigitization of at least some portions of the entireprocess The reengineering exercise made each activ-ity in a process more visible and often certain activi-ties were modularized thereby permitting detailedconsideration of the most cost-effective way ofcompleting each activity and making it possible toseparate them spatially and organizationally shouldthe need arise3

This reengineering was combined with the in-creasing digitization of the objects of service workthat was ongoing throughout the 1990s Digiti-zation was so important because it removed thephysical medium carrying the information thus free-ing it from the constraints of size and weight(Kenney 1997 Curry 1997) Still few service opera-tions can be done only on the computer (the mod-ern form of mundane work in the terminology ofCallaghan Thompson and Warhurst [2001]) Theyusually require some level of face-to-faceinteractivity either among coworkers or with per-sons outside the organizations such as vendors andclients However information technology (IT) and

Volume 1 Number 2 Winter 2003 23

Dossani Kenney

2 We dene a business process as a complete service such as handling a customer complaint processing a medicalclaim or processing a purchase order Completing a process requires undertaking a set of activities For example inhandling a customer complaint it is necessary to understand the complaint classify it decide on a course of action un-dertake the action and follow up to ensure the action solved the complaint Each of these is an activity that is poten-tially separable from the others3 On modularity see for example Baldwin and Clark (2000) and Sturgeon (2002)

Table 1 Sample Firm Size

No of rmsinterviewed

Employmentin sampled

rms

Totalindustry

employment

Sample ofindustry

employment

Averageemployees

per rm

Medianemployees

per rm

42 69729 171500 407 1660 1000

reengineering permitted processes to be parsed intocomponents requiring different levels of skill andinteractivity This made it more feasible to relocatethose activities requiring low levels of face-to-faceinteractivity offshore Orlikowski (1996) captures thisphenomenon in her study of technical help deskswhere digitization induces work ows to be reorga-nized in the direction of dividing work ow intotasks that are separable in terms of technical skillsand interactivity These developments had signicantimplications for the possibility of offshoring even inthe absence of cost arbitrage opportunities that areavailable when work is outsourced to developingnations such as India

The initial relocation of BPs was intranational Forexample call centers were relocated from high-costcities to lower-cost small towns and then to rural lo-cations The earlier efforts to lower wage costs forback-ofce BPs saw rms move their back-ofce op-erations to smaller Midwestern towns where accentswere neutral education was adequate and laborcosts were lower and (at that time) relatively reliableThe cost savings were likely in the 20 to 30range when the low-cost land and labor costs weretotaled However labor pools were shallow thuslimiting the scale of operations

From this perspective the relocation of suchwork to other countries was an extension of thisprocess as rms sought to cut costs further Therewere technical developments that facilitated reloca-tion One of the most important was the continuouscost reduction in data transmission These technicaldevelopments loosened the geographical constraintsthat encouraged the undertaking of BPs nationallyReducing the physical and technical constraintsmeant that variables such as factor costs and skillavailability increasingly became the determinants ofwhere service activities were undertaken

In choosing which processes to undertake off-shore it may be thought that the simplest processeswould tend to be offshored rst because the skillsfor undertaking more complex processes might takea longer time to learn Most existing forms of glob-alization are not processual and thus do not consid-er the evolution of the overseas facility however asKogut and Zander (1992) among others hypothe-size overseas operations also experience learningeven if it is simply catch-up From the perspective ofthe source nation the initial transfer may be onlythe rst stage in the absorption of an entire chain of

activities For example in insurance claimsprocessing whereas initially the Indian operationmight simply process the claim it might be possibleto transfer some of the investigation and valuationactivities to the Indian operation With experienceIndian accountants or engineers may be trained todetermine fraud or exaggeration of claims or at theleast ag unusual claims

There are several additional dimensions to theoffshoring decisions other than the complexity oftasks and the need for face-to-face interactionSome of these are internal to the rm and some arepart of the environment It may happenmdash as wehave noticed in Indiamdashthat highly technical com-plex tasks in customer support that require interac-tion between say a corporate client in the UnitedStates and the vendorrsquos technician may be moreeasily offshored than a simpler task requiring inter-action with a retail customer This is because the lat-ter requires a kind of interactive skill that might beweaker in India compared with the former whereasIndia offers technical skills of high quality capable ofhandling the former task

We identify the following six internal factors thatrms consider before offshoring an activity in addi-tion to environmental factors such as technologicalenablers and the savings in direct labor costs (whichwe consider later)

The knowledge component of the activity Ahigher knowledge component makes the rmmore concerned about whether the quality ofthe service will change because of a locationalchange and greater difculty in the transferprocess There may be reasons to worry aboutquality slippage if the remote location cannotunderstand the quality or qualities needed or(even if it understands) cannot match the qual-ity needed Sometimes if quality is acceptablea rm may not only shift to a remote locationbut may also outsource the activity to mitigatethe possible political difculties of an internaltransfer A rm also might be concerned abouta loss of competencies in a certain locationthat would be costly (or even impossible) to re-acquire in the same location if ever needed Ifunique skills atrophy any disruption of accesscould have dire consequences

The interactive components of the processThere are two dimensions of interactivity that

24 Information Technologies and International Development

Lift and Shift

affect the location of BPs (1) interaction withother persons in the production of the serviceand (2) interaction with the (clientrsquos) customerand other vendors The greater the need to in-teract with other persons within the rm andthe greater the interaction across different ser-vices and processes the higher the cost andrisk threshold of offshoring portions of thework Of course it might be possible to off-shore the entire process thus retaininginteractivity at the new location But if someactivities cannot be offshored offshoring theothers might not be feasible In cases whereoffshoring and outsourcing were both beingconsidered the risks might be even greater Forcertain rms customer interaction is a coreskill If they believe offshoring could disruptsuch interaction they are unlikely to take therisk

The level of separability of the process In cer-tain kinds of work such as customer care theentire process can be offshored because it doesnot require personal interaction These activitiesusually are routinized and even scripted For ex-ample much software support is providedthrough scripted decision trees For lessscripted activities offshoring may be less feasi-ble Consider a process within which 70 ofthe work is Web or database research and30 is interacting directly with clients Unlessit is possible to separate these processes intoactivities it will be impossible to transfer theprocess Separability is not dichotomous butrather a matter of degrees It may be possibleto reengineer a process to make it more sepa-rable though this may be at the cost of somelevel of service

Savings from concentrating an activity in onelocation Often rms will have several ofcesperforming the same function This is often thecase because as rms expand they often tapout local labor pools and therefore establishofces in other regions This can result in aninefcient spatial posture Often in developedcountries it is too expensive to concentratethese facilities in a single location Howeverthe rm relocating to India can overcome thissunk-cost issue The advantages of concentra-tion may be signicant

There are benets of size due to scaling upoperations For example a larger facility ismore capable of operating overload A largerfacility requires a smaller excess cushion tomanage a peak load thereby creating both amanpower and facility savings Of course theremay also be diseconomies of size The advan-tage of concentration can also favor out-sourcing because a service provider can poolthe business of many clients The classic exam-ple is medical transcription which is still mainlyoutsourced to small local rms or even individ-uals in the United States The large rms in In-dia can offer guarantees of quality that smallerdomestic operations cannot offer because theyhave large labor pools thus reducing the ef-fects of absenteeism and offering the typicalefciencies that Adam Smith ([1776] 2003)points out regarding even a simple division oflabor

Reengineering as part of the transfer processTo transfer a business process it is necessary tostudy it intensively and script the transfer Inthe process of study often there will be stepsof the current process that can be discardedconsolidated or simplied These are legaciesof earlier methodologies that were changed asthe production process evolved During thetransfer these are easier to abandon than at anexisting facility where they have become a nat-ural component of the daily routine There areother less expected benets that can arise fromthe transfer process that go beyond theefciency effects These actions of examiningand transferring a process can yield signicantefciencies Processes often evolve in a path-dependent manner and as Arthur (1994) andDavid (1986) show path-dependent outcomesmay not be the most efcient conguration ve-hicle During the transfer process theseinefciencies can be addressed without disrupt-ing work patterns The workers in the new lo-cation are met with a fait accompli

The time-sensitive nature of the work Somework such as payroll has deadlines Indian op-erations can shorten various product cyclesThe most obvious of these is work that needsto be completed overnight in Europe or theUnited States that can be undertaken during

Volume 1 Number 2 Winter 2003 25

Dossani Kenney

the day in India For example technical calls byhigh-end clients may need to be answeredwithin 15 minutes For medical transcriptiondoctors notes for patients in intensive caremight have turnarounds as short as 2 hoursIndian operations are able to undertake thesehigh-priority tasks because they can affordmore slack resources to meet peak loads thantheir Western counterparts can This dimensionof quality is far more expensive to offer in de-veloped nations where labor is more costly

The combination of low labor costs project man-agement skills and technological sophisticationmake India a particularly attractive candidate forBP offshoring4 The genesis of the relocation ofbusiness processes to India can be traced to themid-1980s when India emerged as an offshoresite for software production By the late 1990s In-dia had become a leading supplier of contract soft-ware programming because of its combination ofskilled low-cost manpower and project manage-ment skills

Labor cost differentials are not the only differ-ence In the United States call centers have beenrelocated to low-labor-cost areas which are mostoften rural areas There they are limited by laborshortages high attrition rates and relatively pooreducational levels In contrast Indian cities such asMumbai or Bangalore offer access to enormouslabor pools of potential employees and have thecomplementary urban services that permit the estab-lishment of operations employing thousands Thecall centers in India we visited varied in size but themedian was 1000 employees (see Table 1) The av-erage size in the United States is under 400 andmany are between 150 and 300 employees

The key benets for foreign rms to operate inIndia are cost and labor supply According to an In-dian trade group NASSCOM-McKinsey (2003) BPoffshoring employment for overseas customers was171500 as of March 2003 compared with 106500in March 20025 Revenue generated for the nancial

year ending March 2003 was estimated at $2375billion compared with $1475 billion in the previousyear This implies that revenue per employee was$13848 in 2002ndash 2003 unchanged from the pre-vious year

According to the NASSCOM estimates directcosts per employee are $10354 (or $520 per bill-able hour of which $310 is the estimated laborcost) yielding a 337 return on the direct cost ofeach employee seat Comparably direct costs in theUnited States are estimated by NASSCOM at$55598 per employee (or $2780 per billable hourof which $2150 is the estimated labor cost)6

Clearly the cost pressure to shift the business over-seas especially during the current economic down-turn must be great

Because the median size of rm is 1000 persons(see Table 1) with typical xed-cost investment of$7500 per seat the annual return on capital is466 and average gross prot (which is tax ex-empt under current law) is about $35 million OneFortune 500 rm that consolidated severalfulllment operations to Bangalore reported to usthat the overall cost savings were 807 These rep-resent signicant dollar savings The NASSCOM-McKinsey report (2003) nds that General Electricone of the pioneers of outsourcing service opera-tions to India had achieved an annual savings of$340 million per year from its Indian operationsnow 7 years old Even if these numbers are inatedthe savings are remarkable and accrue directly to thermrsquos protability8

The NASSCOM-McKinsey (2003) report estimatesare that the Information Technology Enabled Ser-vices (ITES) operations will employ between 900000and 1000000 people in 2008 This assumes a com-pound rate of growth of approximately 45 peryear for the next 5 years an assumption yet to bevalidated This is possible because India has thelabor pool the enabling technologies exist and areimproving and the pressure on rms in developednations to lower costs is great If the current trajec-tory continues offshoring may prove to be of greatsignicance to nations and regions that could expe-

26 Information Technologies and International Development

Lift and Shift

4 For an early discussion of the role of India in BP offshoring see Aron and Singh (2002)5 Data for this section draw on NASSCOM-McKinsey (2003) pp 56ndash716 Note that this shows that infrastructure costs are lower in India This was probably always true for most infrastruc-ture though the plummeting cost of data communications (PCs software and telecommunications) to global levels isa recent outcome of reform as is the low cost of nance7 Personal interview by authors April 20038 The authorsrsquo interviews appear to validate the savings potential

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

operations as small as 100 employees Hence webelieve that the rms we interviewed are representa-tive of the Indian BP offshoring industry

The interviewees were executives usually vicepresidents with various functional titles and infewer cases the CEOs or chairpersons The inter-views were invariably at least 1 hour in length andoften up to 2 hours In 70 of the cases there wasa tour of the facilities though because of the timedifference with the United States the facilities usu-ally had minimal activity All interviews were tapedand transcribed Unfortunately because of the polit-ical and commercial sensitivity of business processoffshoring the respondents requested anonymity

In addition to the interviews we reviewed the lit-erature on BP offshoring and outsourcing This liter-ature falls in two realms consultant reports and thepopular press The consultant reports can be valu-able but frequently suffer from their desire to createbusiness for themselves The press reports are usefulfor providing topical materials especially in follow-ing investments and expansions However exceptfor these two sources of written material there is lit-tle other material on the topic

Business processes is the catchall term used for themyriad white-collar processes that any bureaucraticentity undertakes in servicing its employees ven-dors and customers2 These include human re-sources accounting auditing customer caretelemarketing tax preparation claims processingdocument management and many other chores

necessary for rm functioning Invariably the com-pletion of a BP is the result of an entire chain ofseparate activities

Until the early 1990s business processes weregenerally treated as a xed cost and received mini-mal management attention The interest inreengineering awakened management attention tothe savings that could be achieved by reorganizingthese processes (Hammer 1990 Hammer andChampy 1993 Davenport 1993 Cole 1994) Onepart counseled the decomposition examination andstandardization of the activities necessary to com-plete a process This was often accompanied by adigitization of at least some portions of the entireprocess The reengineering exercise made each activ-ity in a process more visible and often certain activi-ties were modularized thereby permitting detailedconsideration of the most cost-effective way ofcompleting each activity and making it possible toseparate them spatially and organizationally shouldthe need arise3

This reengineering was combined with the in-creasing digitization of the objects of service workthat was ongoing throughout the 1990s Digiti-zation was so important because it removed thephysical medium carrying the information thus free-ing it from the constraints of size and weight(Kenney 1997 Curry 1997) Still few service opera-tions can be done only on the computer (the mod-ern form of mundane work in the terminology ofCallaghan Thompson and Warhurst [2001]) Theyusually require some level of face-to-faceinteractivity either among coworkers or with per-sons outside the organizations such as vendors andclients However information technology (IT) and

Volume 1 Number 2 Winter 2003 23

Dossani Kenney

2 We dene a business process as a complete service such as handling a customer complaint processing a medicalclaim or processing a purchase order Completing a process requires undertaking a set of activities For example inhandling a customer complaint it is necessary to understand the complaint classify it decide on a course of action un-dertake the action and follow up to ensure the action solved the complaint Each of these is an activity that is poten-tially separable from the others3 On modularity see for example Baldwin and Clark (2000) and Sturgeon (2002)

Table 1 Sample Firm Size

No of rmsinterviewed

Employmentin sampled

rms

Totalindustry

employment

Sample ofindustry

employment

Averageemployees

per rm

Medianemployees

per rm

42 69729 171500 407 1660 1000

reengineering permitted processes to be parsed intocomponents requiring different levels of skill andinteractivity This made it more feasible to relocatethose activities requiring low levels of face-to-faceinteractivity offshore Orlikowski (1996) captures thisphenomenon in her study of technical help deskswhere digitization induces work ows to be reorga-nized in the direction of dividing work ow intotasks that are separable in terms of technical skillsand interactivity These developments had signicantimplications for the possibility of offshoring even inthe absence of cost arbitrage opportunities that areavailable when work is outsourced to developingnations such as India

The initial relocation of BPs was intranational Forexample call centers were relocated from high-costcities to lower-cost small towns and then to rural lo-cations The earlier efforts to lower wage costs forback-ofce BPs saw rms move their back-ofce op-erations to smaller Midwestern towns where accentswere neutral education was adequate and laborcosts were lower and (at that time) relatively reliableThe cost savings were likely in the 20 to 30range when the low-cost land and labor costs weretotaled However labor pools were shallow thuslimiting the scale of operations

From this perspective the relocation of suchwork to other countries was an extension of thisprocess as rms sought to cut costs further Therewere technical developments that facilitated reloca-tion One of the most important was the continuouscost reduction in data transmission These technicaldevelopments loosened the geographical constraintsthat encouraged the undertaking of BPs nationallyReducing the physical and technical constraintsmeant that variables such as factor costs and skillavailability increasingly became the determinants ofwhere service activities were undertaken

In choosing which processes to undertake off-shore it may be thought that the simplest processeswould tend to be offshored rst because the skillsfor undertaking more complex processes might takea longer time to learn Most existing forms of glob-alization are not processual and thus do not consid-er the evolution of the overseas facility however asKogut and Zander (1992) among others hypothe-size overseas operations also experience learningeven if it is simply catch-up From the perspective ofthe source nation the initial transfer may be onlythe rst stage in the absorption of an entire chain of

activities For example in insurance claimsprocessing whereas initially the Indian operationmight simply process the claim it might be possibleto transfer some of the investigation and valuationactivities to the Indian operation With experienceIndian accountants or engineers may be trained todetermine fraud or exaggeration of claims or at theleast ag unusual claims

There are several additional dimensions to theoffshoring decisions other than the complexity oftasks and the need for face-to-face interactionSome of these are internal to the rm and some arepart of the environment It may happenmdash as wehave noticed in Indiamdashthat highly technical com-plex tasks in customer support that require interac-tion between say a corporate client in the UnitedStates and the vendorrsquos technician may be moreeasily offshored than a simpler task requiring inter-action with a retail customer This is because the lat-ter requires a kind of interactive skill that might beweaker in India compared with the former whereasIndia offers technical skills of high quality capable ofhandling the former task

We identify the following six internal factors thatrms consider before offshoring an activity in addi-tion to environmental factors such as technologicalenablers and the savings in direct labor costs (whichwe consider later)

The knowledge component of the activity Ahigher knowledge component makes the rmmore concerned about whether the quality ofthe service will change because of a locationalchange and greater difculty in the transferprocess There may be reasons to worry aboutquality slippage if the remote location cannotunderstand the quality or qualities needed or(even if it understands) cannot match the qual-ity needed Sometimes if quality is acceptablea rm may not only shift to a remote locationbut may also outsource the activity to mitigatethe possible political difculties of an internaltransfer A rm also might be concerned abouta loss of competencies in a certain locationthat would be costly (or even impossible) to re-acquire in the same location if ever needed Ifunique skills atrophy any disruption of accesscould have dire consequences

The interactive components of the processThere are two dimensions of interactivity that

24 Information Technologies and International Development

Lift and Shift

affect the location of BPs (1) interaction withother persons in the production of the serviceand (2) interaction with the (clientrsquos) customerand other vendors The greater the need to in-teract with other persons within the rm andthe greater the interaction across different ser-vices and processes the higher the cost andrisk threshold of offshoring portions of thework Of course it might be possible to off-shore the entire process thus retaininginteractivity at the new location But if someactivities cannot be offshored offshoring theothers might not be feasible In cases whereoffshoring and outsourcing were both beingconsidered the risks might be even greater Forcertain rms customer interaction is a coreskill If they believe offshoring could disruptsuch interaction they are unlikely to take therisk

The level of separability of the process In cer-tain kinds of work such as customer care theentire process can be offshored because it doesnot require personal interaction These activitiesusually are routinized and even scripted For ex-ample much software support is providedthrough scripted decision trees For lessscripted activities offshoring may be less feasi-ble Consider a process within which 70 ofthe work is Web or database research and30 is interacting directly with clients Unlessit is possible to separate these processes intoactivities it will be impossible to transfer theprocess Separability is not dichotomous butrather a matter of degrees It may be possibleto reengineer a process to make it more sepa-rable though this may be at the cost of somelevel of service

Savings from concentrating an activity in onelocation Often rms will have several ofcesperforming the same function This is often thecase because as rms expand they often tapout local labor pools and therefore establishofces in other regions This can result in aninefcient spatial posture Often in developedcountries it is too expensive to concentratethese facilities in a single location Howeverthe rm relocating to India can overcome thissunk-cost issue The advantages of concentra-tion may be signicant

There are benets of size due to scaling upoperations For example a larger facility ismore capable of operating overload A largerfacility requires a smaller excess cushion tomanage a peak load thereby creating both amanpower and facility savings Of course theremay also be diseconomies of size The advan-tage of concentration can also favor out-sourcing because a service provider can poolthe business of many clients The classic exam-ple is medical transcription which is still mainlyoutsourced to small local rms or even individ-uals in the United States The large rms in In-dia can offer guarantees of quality that smallerdomestic operations cannot offer because theyhave large labor pools thus reducing the ef-fects of absenteeism and offering the typicalefciencies that Adam Smith ([1776] 2003)points out regarding even a simple division oflabor

Reengineering as part of the transfer processTo transfer a business process it is necessary tostudy it intensively and script the transfer Inthe process of study often there will be stepsof the current process that can be discardedconsolidated or simplied These are legaciesof earlier methodologies that were changed asthe production process evolved During thetransfer these are easier to abandon than at anexisting facility where they have become a nat-ural component of the daily routine There areother less expected benets that can arise fromthe transfer process that go beyond theefciency effects These actions of examiningand transferring a process can yield signicantefciencies Processes often evolve in a path-dependent manner and as Arthur (1994) andDavid (1986) show path-dependent outcomesmay not be the most efcient conguration ve-hicle During the transfer process theseinefciencies can be addressed without disrupt-ing work patterns The workers in the new lo-cation are met with a fait accompli

The time-sensitive nature of the work Somework such as payroll has deadlines Indian op-erations can shorten various product cyclesThe most obvious of these is work that needsto be completed overnight in Europe or theUnited States that can be undertaken during

Volume 1 Number 2 Winter 2003 25

Dossani Kenney

the day in India For example technical calls byhigh-end clients may need to be answeredwithin 15 minutes For medical transcriptiondoctors notes for patients in intensive caremight have turnarounds as short as 2 hoursIndian operations are able to undertake thesehigh-priority tasks because they can affordmore slack resources to meet peak loads thantheir Western counterparts can This dimensionof quality is far more expensive to offer in de-veloped nations where labor is more costly

The combination of low labor costs project man-agement skills and technological sophisticationmake India a particularly attractive candidate forBP offshoring4 The genesis of the relocation ofbusiness processes to India can be traced to themid-1980s when India emerged as an offshoresite for software production By the late 1990s In-dia had become a leading supplier of contract soft-ware programming because of its combination ofskilled low-cost manpower and project manage-ment skills

Labor cost differentials are not the only differ-ence In the United States call centers have beenrelocated to low-labor-cost areas which are mostoften rural areas There they are limited by laborshortages high attrition rates and relatively pooreducational levels In contrast Indian cities such asMumbai or Bangalore offer access to enormouslabor pools of potential employees and have thecomplementary urban services that permit the estab-lishment of operations employing thousands Thecall centers in India we visited varied in size but themedian was 1000 employees (see Table 1) The av-erage size in the United States is under 400 andmany are between 150 and 300 employees

The key benets for foreign rms to operate inIndia are cost and labor supply According to an In-dian trade group NASSCOM-McKinsey (2003) BPoffshoring employment for overseas customers was171500 as of March 2003 compared with 106500in March 20025 Revenue generated for the nancial

year ending March 2003 was estimated at $2375billion compared with $1475 billion in the previousyear This implies that revenue per employee was$13848 in 2002ndash 2003 unchanged from the pre-vious year

According to the NASSCOM estimates directcosts per employee are $10354 (or $520 per bill-able hour of which $310 is the estimated laborcost) yielding a 337 return on the direct cost ofeach employee seat Comparably direct costs in theUnited States are estimated by NASSCOM at$55598 per employee (or $2780 per billable hourof which $2150 is the estimated labor cost)6

Clearly the cost pressure to shift the business over-seas especially during the current economic down-turn must be great

Because the median size of rm is 1000 persons(see Table 1) with typical xed-cost investment of$7500 per seat the annual return on capital is466 and average gross prot (which is tax ex-empt under current law) is about $35 million OneFortune 500 rm that consolidated severalfulllment operations to Bangalore reported to usthat the overall cost savings were 807 These rep-resent signicant dollar savings The NASSCOM-McKinsey report (2003) nds that General Electricone of the pioneers of outsourcing service opera-tions to India had achieved an annual savings of$340 million per year from its Indian operationsnow 7 years old Even if these numbers are inatedthe savings are remarkable and accrue directly to thermrsquos protability8

The NASSCOM-McKinsey (2003) report estimatesare that the Information Technology Enabled Ser-vices (ITES) operations will employ between 900000and 1000000 people in 2008 This assumes a com-pound rate of growth of approximately 45 peryear for the next 5 years an assumption yet to bevalidated This is possible because India has thelabor pool the enabling technologies exist and areimproving and the pressure on rms in developednations to lower costs is great If the current trajec-tory continues offshoring may prove to be of greatsignicance to nations and regions that could expe-

26 Information Technologies and International Development

Lift and Shift

4 For an early discussion of the role of India in BP offshoring see Aron and Singh (2002)5 Data for this section draw on NASSCOM-McKinsey (2003) pp 56ndash716 Note that this shows that infrastructure costs are lower in India This was probably always true for most infrastruc-ture though the plummeting cost of data communications (PCs software and telecommunications) to global levels isa recent outcome of reform as is the low cost of nance7 Personal interview by authors April 20038 The authorsrsquo interviews appear to validate the savings potential

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

reengineering permitted processes to be parsed intocomponents requiring different levels of skill andinteractivity This made it more feasible to relocatethose activities requiring low levels of face-to-faceinteractivity offshore Orlikowski (1996) captures thisphenomenon in her study of technical help deskswhere digitization induces work ows to be reorga-nized in the direction of dividing work ow intotasks that are separable in terms of technical skillsand interactivity These developments had signicantimplications for the possibility of offshoring even inthe absence of cost arbitrage opportunities that areavailable when work is outsourced to developingnations such as India

The initial relocation of BPs was intranational Forexample call centers were relocated from high-costcities to lower-cost small towns and then to rural lo-cations The earlier efforts to lower wage costs forback-ofce BPs saw rms move their back-ofce op-erations to smaller Midwestern towns where accentswere neutral education was adequate and laborcosts were lower and (at that time) relatively reliableThe cost savings were likely in the 20 to 30range when the low-cost land and labor costs weretotaled However labor pools were shallow thuslimiting the scale of operations

From this perspective the relocation of suchwork to other countries was an extension of thisprocess as rms sought to cut costs further Therewere technical developments that facilitated reloca-tion One of the most important was the continuouscost reduction in data transmission These technicaldevelopments loosened the geographical constraintsthat encouraged the undertaking of BPs nationallyReducing the physical and technical constraintsmeant that variables such as factor costs and skillavailability increasingly became the determinants ofwhere service activities were undertaken

In choosing which processes to undertake off-shore it may be thought that the simplest processeswould tend to be offshored rst because the skillsfor undertaking more complex processes might takea longer time to learn Most existing forms of glob-alization are not processual and thus do not consid-er the evolution of the overseas facility however asKogut and Zander (1992) among others hypothe-size overseas operations also experience learningeven if it is simply catch-up From the perspective ofthe source nation the initial transfer may be onlythe rst stage in the absorption of an entire chain of

activities For example in insurance claimsprocessing whereas initially the Indian operationmight simply process the claim it might be possibleto transfer some of the investigation and valuationactivities to the Indian operation With experienceIndian accountants or engineers may be trained todetermine fraud or exaggeration of claims or at theleast ag unusual claims

There are several additional dimensions to theoffshoring decisions other than the complexity oftasks and the need for face-to-face interactionSome of these are internal to the rm and some arepart of the environment It may happenmdash as wehave noticed in Indiamdashthat highly technical com-plex tasks in customer support that require interac-tion between say a corporate client in the UnitedStates and the vendorrsquos technician may be moreeasily offshored than a simpler task requiring inter-action with a retail customer This is because the lat-ter requires a kind of interactive skill that might beweaker in India compared with the former whereasIndia offers technical skills of high quality capable ofhandling the former task

We identify the following six internal factors thatrms consider before offshoring an activity in addi-tion to environmental factors such as technologicalenablers and the savings in direct labor costs (whichwe consider later)

The knowledge component of the activity Ahigher knowledge component makes the rmmore concerned about whether the quality ofthe service will change because of a locationalchange and greater difculty in the transferprocess There may be reasons to worry aboutquality slippage if the remote location cannotunderstand the quality or qualities needed or(even if it understands) cannot match the qual-ity needed Sometimes if quality is acceptablea rm may not only shift to a remote locationbut may also outsource the activity to mitigatethe possible political difculties of an internaltransfer A rm also might be concerned abouta loss of competencies in a certain locationthat would be costly (or even impossible) to re-acquire in the same location if ever needed Ifunique skills atrophy any disruption of accesscould have dire consequences

The interactive components of the processThere are two dimensions of interactivity that

24 Information Technologies and International Development

Lift and Shift

affect the location of BPs (1) interaction withother persons in the production of the serviceand (2) interaction with the (clientrsquos) customerand other vendors The greater the need to in-teract with other persons within the rm andthe greater the interaction across different ser-vices and processes the higher the cost andrisk threshold of offshoring portions of thework Of course it might be possible to off-shore the entire process thus retaininginteractivity at the new location But if someactivities cannot be offshored offshoring theothers might not be feasible In cases whereoffshoring and outsourcing were both beingconsidered the risks might be even greater Forcertain rms customer interaction is a coreskill If they believe offshoring could disruptsuch interaction they are unlikely to take therisk

The level of separability of the process In cer-tain kinds of work such as customer care theentire process can be offshored because it doesnot require personal interaction These activitiesusually are routinized and even scripted For ex-ample much software support is providedthrough scripted decision trees For lessscripted activities offshoring may be less feasi-ble Consider a process within which 70 ofthe work is Web or database research and30 is interacting directly with clients Unlessit is possible to separate these processes intoactivities it will be impossible to transfer theprocess Separability is not dichotomous butrather a matter of degrees It may be possibleto reengineer a process to make it more sepa-rable though this may be at the cost of somelevel of service

Savings from concentrating an activity in onelocation Often rms will have several ofcesperforming the same function This is often thecase because as rms expand they often tapout local labor pools and therefore establishofces in other regions This can result in aninefcient spatial posture Often in developedcountries it is too expensive to concentratethese facilities in a single location Howeverthe rm relocating to India can overcome thissunk-cost issue The advantages of concentra-tion may be signicant

There are benets of size due to scaling upoperations For example a larger facility ismore capable of operating overload A largerfacility requires a smaller excess cushion tomanage a peak load thereby creating both amanpower and facility savings Of course theremay also be diseconomies of size The advan-tage of concentration can also favor out-sourcing because a service provider can poolthe business of many clients The classic exam-ple is medical transcription which is still mainlyoutsourced to small local rms or even individ-uals in the United States The large rms in In-dia can offer guarantees of quality that smallerdomestic operations cannot offer because theyhave large labor pools thus reducing the ef-fects of absenteeism and offering the typicalefciencies that Adam Smith ([1776] 2003)points out regarding even a simple division oflabor

Reengineering as part of the transfer processTo transfer a business process it is necessary tostudy it intensively and script the transfer Inthe process of study often there will be stepsof the current process that can be discardedconsolidated or simplied These are legaciesof earlier methodologies that were changed asthe production process evolved During thetransfer these are easier to abandon than at anexisting facility where they have become a nat-ural component of the daily routine There areother less expected benets that can arise fromthe transfer process that go beyond theefciency effects These actions of examiningand transferring a process can yield signicantefciencies Processes often evolve in a path-dependent manner and as Arthur (1994) andDavid (1986) show path-dependent outcomesmay not be the most efcient conguration ve-hicle During the transfer process theseinefciencies can be addressed without disrupt-ing work patterns The workers in the new lo-cation are met with a fait accompli

The time-sensitive nature of the work Somework such as payroll has deadlines Indian op-erations can shorten various product cyclesThe most obvious of these is work that needsto be completed overnight in Europe or theUnited States that can be undertaken during

Volume 1 Number 2 Winter 2003 25

Dossani Kenney

the day in India For example technical calls byhigh-end clients may need to be answeredwithin 15 minutes For medical transcriptiondoctors notes for patients in intensive caremight have turnarounds as short as 2 hoursIndian operations are able to undertake thesehigh-priority tasks because they can affordmore slack resources to meet peak loads thantheir Western counterparts can This dimensionof quality is far more expensive to offer in de-veloped nations where labor is more costly

The combination of low labor costs project man-agement skills and technological sophisticationmake India a particularly attractive candidate forBP offshoring4 The genesis of the relocation ofbusiness processes to India can be traced to themid-1980s when India emerged as an offshoresite for software production By the late 1990s In-dia had become a leading supplier of contract soft-ware programming because of its combination ofskilled low-cost manpower and project manage-ment skills

Labor cost differentials are not the only differ-ence In the United States call centers have beenrelocated to low-labor-cost areas which are mostoften rural areas There they are limited by laborshortages high attrition rates and relatively pooreducational levels In contrast Indian cities such asMumbai or Bangalore offer access to enormouslabor pools of potential employees and have thecomplementary urban services that permit the estab-lishment of operations employing thousands Thecall centers in India we visited varied in size but themedian was 1000 employees (see Table 1) The av-erage size in the United States is under 400 andmany are between 150 and 300 employees

The key benets for foreign rms to operate inIndia are cost and labor supply According to an In-dian trade group NASSCOM-McKinsey (2003) BPoffshoring employment for overseas customers was171500 as of March 2003 compared with 106500in March 20025 Revenue generated for the nancial

year ending March 2003 was estimated at $2375billion compared with $1475 billion in the previousyear This implies that revenue per employee was$13848 in 2002ndash 2003 unchanged from the pre-vious year

According to the NASSCOM estimates directcosts per employee are $10354 (or $520 per bill-able hour of which $310 is the estimated laborcost) yielding a 337 return on the direct cost ofeach employee seat Comparably direct costs in theUnited States are estimated by NASSCOM at$55598 per employee (or $2780 per billable hourof which $2150 is the estimated labor cost)6

Clearly the cost pressure to shift the business over-seas especially during the current economic down-turn must be great

Because the median size of rm is 1000 persons(see Table 1) with typical xed-cost investment of$7500 per seat the annual return on capital is466 and average gross prot (which is tax ex-empt under current law) is about $35 million OneFortune 500 rm that consolidated severalfulllment operations to Bangalore reported to usthat the overall cost savings were 807 These rep-resent signicant dollar savings The NASSCOM-McKinsey report (2003) nds that General Electricone of the pioneers of outsourcing service opera-tions to India had achieved an annual savings of$340 million per year from its Indian operationsnow 7 years old Even if these numbers are inatedthe savings are remarkable and accrue directly to thermrsquos protability8

The NASSCOM-McKinsey (2003) report estimatesare that the Information Technology Enabled Ser-vices (ITES) operations will employ between 900000and 1000000 people in 2008 This assumes a com-pound rate of growth of approximately 45 peryear for the next 5 years an assumption yet to bevalidated This is possible because India has thelabor pool the enabling technologies exist and areimproving and the pressure on rms in developednations to lower costs is great If the current trajec-tory continues offshoring may prove to be of greatsignicance to nations and regions that could expe-

26 Information Technologies and International Development

Lift and Shift

4 For an early discussion of the role of India in BP offshoring see Aron and Singh (2002)5 Data for this section draw on NASSCOM-McKinsey (2003) pp 56ndash716 Note that this shows that infrastructure costs are lower in India This was probably always true for most infrastruc-ture though the plummeting cost of data communications (PCs software and telecommunications) to global levels isa recent outcome of reform as is the low cost of nance7 Personal interview by authors April 20038 The authorsrsquo interviews appear to validate the savings potential

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

affect the location of BPs (1) interaction withother persons in the production of the serviceand (2) interaction with the (clientrsquos) customerand other vendors The greater the need to in-teract with other persons within the rm andthe greater the interaction across different ser-vices and processes the higher the cost andrisk threshold of offshoring portions of thework Of course it might be possible to off-shore the entire process thus retaininginteractivity at the new location But if someactivities cannot be offshored offshoring theothers might not be feasible In cases whereoffshoring and outsourcing were both beingconsidered the risks might be even greater Forcertain rms customer interaction is a coreskill If they believe offshoring could disruptsuch interaction they are unlikely to take therisk

The level of separability of the process In cer-tain kinds of work such as customer care theentire process can be offshored because it doesnot require personal interaction These activitiesusually are routinized and even scripted For ex-ample much software support is providedthrough scripted decision trees For lessscripted activities offshoring may be less feasi-ble Consider a process within which 70 ofthe work is Web or database research and30 is interacting directly with clients Unlessit is possible to separate these processes intoactivities it will be impossible to transfer theprocess Separability is not dichotomous butrather a matter of degrees It may be possibleto reengineer a process to make it more sepa-rable though this may be at the cost of somelevel of service

Savings from concentrating an activity in onelocation Often rms will have several ofcesperforming the same function This is often thecase because as rms expand they often tapout local labor pools and therefore establishofces in other regions This can result in aninefcient spatial posture Often in developedcountries it is too expensive to concentratethese facilities in a single location Howeverthe rm relocating to India can overcome thissunk-cost issue The advantages of concentra-tion may be signicant

There are benets of size due to scaling upoperations For example a larger facility ismore capable of operating overload A largerfacility requires a smaller excess cushion tomanage a peak load thereby creating both amanpower and facility savings Of course theremay also be diseconomies of size The advan-tage of concentration can also favor out-sourcing because a service provider can poolthe business of many clients The classic exam-ple is medical transcription which is still mainlyoutsourced to small local rms or even individ-uals in the United States The large rms in In-dia can offer guarantees of quality that smallerdomestic operations cannot offer because theyhave large labor pools thus reducing the ef-fects of absenteeism and offering the typicalefciencies that Adam Smith ([1776] 2003)points out regarding even a simple division oflabor

Reengineering as part of the transfer processTo transfer a business process it is necessary tostudy it intensively and script the transfer Inthe process of study often there will be stepsof the current process that can be discardedconsolidated or simplied These are legaciesof earlier methodologies that were changed asthe production process evolved During thetransfer these are easier to abandon than at anexisting facility where they have become a nat-ural component of the daily routine There areother less expected benets that can arise fromthe transfer process that go beyond theefciency effects These actions of examiningand transferring a process can yield signicantefciencies Processes often evolve in a path-dependent manner and as Arthur (1994) andDavid (1986) show path-dependent outcomesmay not be the most efcient conguration ve-hicle During the transfer process theseinefciencies can be addressed without disrupt-ing work patterns The workers in the new lo-cation are met with a fait accompli

The time-sensitive nature of the work Somework such as payroll has deadlines Indian op-erations can shorten various product cyclesThe most obvious of these is work that needsto be completed overnight in Europe or theUnited States that can be undertaken during

Volume 1 Number 2 Winter 2003 25

Dossani Kenney

the day in India For example technical calls byhigh-end clients may need to be answeredwithin 15 minutes For medical transcriptiondoctors notes for patients in intensive caremight have turnarounds as short as 2 hoursIndian operations are able to undertake thesehigh-priority tasks because they can affordmore slack resources to meet peak loads thantheir Western counterparts can This dimensionof quality is far more expensive to offer in de-veloped nations where labor is more costly

The combination of low labor costs project man-agement skills and technological sophisticationmake India a particularly attractive candidate forBP offshoring4 The genesis of the relocation ofbusiness processes to India can be traced to themid-1980s when India emerged as an offshoresite for software production By the late 1990s In-dia had become a leading supplier of contract soft-ware programming because of its combination ofskilled low-cost manpower and project manage-ment skills

Labor cost differentials are not the only differ-ence In the United States call centers have beenrelocated to low-labor-cost areas which are mostoften rural areas There they are limited by laborshortages high attrition rates and relatively pooreducational levels In contrast Indian cities such asMumbai or Bangalore offer access to enormouslabor pools of potential employees and have thecomplementary urban services that permit the estab-lishment of operations employing thousands Thecall centers in India we visited varied in size but themedian was 1000 employees (see Table 1) The av-erage size in the United States is under 400 andmany are between 150 and 300 employees

The key benets for foreign rms to operate inIndia are cost and labor supply According to an In-dian trade group NASSCOM-McKinsey (2003) BPoffshoring employment for overseas customers was171500 as of March 2003 compared with 106500in March 20025 Revenue generated for the nancial

year ending March 2003 was estimated at $2375billion compared with $1475 billion in the previousyear This implies that revenue per employee was$13848 in 2002ndash 2003 unchanged from the pre-vious year

According to the NASSCOM estimates directcosts per employee are $10354 (or $520 per bill-able hour of which $310 is the estimated laborcost) yielding a 337 return on the direct cost ofeach employee seat Comparably direct costs in theUnited States are estimated by NASSCOM at$55598 per employee (or $2780 per billable hourof which $2150 is the estimated labor cost)6

Clearly the cost pressure to shift the business over-seas especially during the current economic down-turn must be great

Because the median size of rm is 1000 persons(see Table 1) with typical xed-cost investment of$7500 per seat the annual return on capital is466 and average gross prot (which is tax ex-empt under current law) is about $35 million OneFortune 500 rm that consolidated severalfulllment operations to Bangalore reported to usthat the overall cost savings were 807 These rep-resent signicant dollar savings The NASSCOM-McKinsey report (2003) nds that General Electricone of the pioneers of outsourcing service opera-tions to India had achieved an annual savings of$340 million per year from its Indian operationsnow 7 years old Even if these numbers are inatedthe savings are remarkable and accrue directly to thermrsquos protability8

The NASSCOM-McKinsey (2003) report estimatesare that the Information Technology Enabled Ser-vices (ITES) operations will employ between 900000and 1000000 people in 2008 This assumes a com-pound rate of growth of approximately 45 peryear for the next 5 years an assumption yet to bevalidated This is possible because India has thelabor pool the enabling technologies exist and areimproving and the pressure on rms in developednations to lower costs is great If the current trajec-tory continues offshoring may prove to be of greatsignicance to nations and regions that could expe-

26 Information Technologies and International Development

Lift and Shift

4 For an early discussion of the role of India in BP offshoring see Aron and Singh (2002)5 Data for this section draw on NASSCOM-McKinsey (2003) pp 56ndash716 Note that this shows that infrastructure costs are lower in India This was probably always true for most infrastruc-ture though the plummeting cost of data communications (PCs software and telecommunications) to global levels isa recent outcome of reform as is the low cost of nance7 Personal interview by authors April 20038 The authorsrsquo interviews appear to validate the savings potential

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

the day in India For example technical calls byhigh-end clients may need to be answeredwithin 15 minutes For medical transcriptiondoctors notes for patients in intensive caremight have turnarounds as short as 2 hoursIndian operations are able to undertake thesehigh-priority tasks because they can affordmore slack resources to meet peak loads thantheir Western counterparts can This dimensionof quality is far more expensive to offer in de-veloped nations where labor is more costly

The combination of low labor costs project man-agement skills and technological sophisticationmake India a particularly attractive candidate forBP offshoring4 The genesis of the relocation ofbusiness processes to India can be traced to themid-1980s when India emerged as an offshoresite for software production By the late 1990s In-dia had become a leading supplier of contract soft-ware programming because of its combination ofskilled low-cost manpower and project manage-ment skills

Labor cost differentials are not the only differ-ence In the United States call centers have beenrelocated to low-labor-cost areas which are mostoften rural areas There they are limited by laborshortages high attrition rates and relatively pooreducational levels In contrast Indian cities such asMumbai or Bangalore offer access to enormouslabor pools of potential employees and have thecomplementary urban services that permit the estab-lishment of operations employing thousands Thecall centers in India we visited varied in size but themedian was 1000 employees (see Table 1) The av-erage size in the United States is under 400 andmany are between 150 and 300 employees

The key benets for foreign rms to operate inIndia are cost and labor supply According to an In-dian trade group NASSCOM-McKinsey (2003) BPoffshoring employment for overseas customers was171500 as of March 2003 compared with 106500in March 20025 Revenue generated for the nancial

year ending March 2003 was estimated at $2375billion compared with $1475 billion in the previousyear This implies that revenue per employee was$13848 in 2002ndash 2003 unchanged from the pre-vious year

According to the NASSCOM estimates directcosts per employee are $10354 (or $520 per bill-able hour of which $310 is the estimated laborcost) yielding a 337 return on the direct cost ofeach employee seat Comparably direct costs in theUnited States are estimated by NASSCOM at$55598 per employee (or $2780 per billable hourof which $2150 is the estimated labor cost)6

Clearly the cost pressure to shift the business over-seas especially during the current economic down-turn must be great

Because the median size of rm is 1000 persons(see Table 1) with typical xed-cost investment of$7500 per seat the annual return on capital is466 and average gross prot (which is tax ex-empt under current law) is about $35 million OneFortune 500 rm that consolidated severalfulllment operations to Bangalore reported to usthat the overall cost savings were 807 These rep-resent signicant dollar savings The NASSCOM-McKinsey report (2003) nds that General Electricone of the pioneers of outsourcing service opera-tions to India had achieved an annual savings of$340 million per year from its Indian operationsnow 7 years old Even if these numbers are inatedthe savings are remarkable and accrue directly to thermrsquos protability8

The NASSCOM-McKinsey (2003) report estimatesare that the Information Technology Enabled Ser-vices (ITES) operations will employ between 900000and 1000000 people in 2008 This assumes a com-pound rate of growth of approximately 45 peryear for the next 5 years an assumption yet to bevalidated This is possible because India has thelabor pool the enabling technologies exist and areimproving and the pressure on rms in developednations to lower costs is great If the current trajec-tory continues offshoring may prove to be of greatsignicance to nations and regions that could expe-

26 Information Technologies and International Development

Lift and Shift

4 For an early discussion of the role of India in BP offshoring see Aron and Singh (2002)5 Data for this section draw on NASSCOM-McKinsey (2003) pp 56ndash716 Note that this shows that infrastructure costs are lower in India This was probably always true for most infrastruc-ture though the plummeting cost of data communications (PCs software and telecommunications) to global levels isa recent outcome of reform as is the low cost of nance7 Personal interview by authors April 20038 The authorsrsquo interviews appear to validate the savings potential

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

rience substantial job losses as service jobs arerelocated

The rms that have most aggressively offshoredwork to India have been in the health care bankingand insurance sectors The BPs have been medicaltranscription call centers accounting and claimsprocessing The initial activities have been highlyroutinized and resemble the initial phase of softwareoutsourcing where during the rst phase ofoutsourcing the work consisted of highly speciedand routine coding operations9 More complex pro-cesses such as preparation of receivables statementsand managing collections have also proven to beamenable to transfer The next phase may take sev-eral directions First processes linking the organiza-tion with customers or suppliers or supportingproduction processes may be amenable to remotefulllment For example one rm reported that afterbeginning with answering calls from potential clientsfor loan services they went on to prequalifyingloans before transferring them to loan ofcers lo-cated in the United States Another rm had movedfrom medical transcription to coding the transcribedwork into a billable event Also supply chain man-agement and customer care are possible candidatesIn health care clinical trials gene testing and algo-rithm development might be offshored Second asBP outsourcing providers develop expertise throughworking for several clients they may be able tomove upstream and provide advice on business pro-cess reengineering

The longer term drivers of the business in Indiahave been the large-scale reform in the communica-tions infrastructure (Dossani 2002) and to an ex-tent venture capital (Dossani and Kenney 2002)Beginning in 1999 India reformed its public monop-oly telecommunications system into a market-drivensystem through allowing a large number of privateproviders to enter the business The private provid-ers were allowed to choose their specializationsranging from providing niche services such as back-bone and network management to full-service inte-grated voice and data operations The result hasbeen the provisioning of a telecommunications net-

work with quality and cost levels approaching thatof developed countries though mainly in the largercities

Also in 1999 venture capital rules were changedto allow foreign venture capital to invest in Indianstart-ups on terms similar to those available in devel-oping countries This has led to the availability ofnancial support for a number of independent rmsproviding BP offshoring services (see the followingdiscussion of the industry structure)

India faces two key challenges in maintaining thepace of growth although these are not likely tohave a short-term impact The rst is a shortage ofmanagerial talent Such talent is needed for severalactivities many of which are new to India The rstis managing the migration of a business processfrom an overseas rm to the Indian operations Thelarger and apparently more successful rms reportedthat it often took up to a year to make such a trans-fer for some of the more complex back-ofce opera-tions whereas the simpler operations such asoutbound call centers could be transferred within amonth The second managerial task is to maintain aseamless relationship between the transferred entityand the operations in the developed country Thethird managerial task is raising and then maintainingthe productivity of operator-level staff Althoughsome rms notably MNCs were achieving produc-tivity rates that matched or even exceeded those oftheir developed country counterparts this appearsto be a greater problem for independent rms Thishas also been greatly exacerbated by high staff turn-over levels10 Although it is claimed that turnoverrates are lower than in developed countries whichcan be in excess of 100 some Indian providersreported attrition rates of 7 per month although35 per month was the average rate reported inour interviews

The second key challenge for India-based rmsis the shortage of domain (sectoral) expertiseespecially in the fastest-growing BP offshoring in-dustries nance insurance real estate health careand logistics11 Given that India has only recentlyliberalized its foreign investment regulations

Volume 1 Number 2 Winter 2003 27

Dossani Kenney

9 For example much of Y2K rectication work consisted of searching for and expanding date elds from two to fourdigits in the source code10 Developed countries also have high turnover rates probably because of the relatively low wage rate and highstress levels In India our interviewees indicated that they viewed the work as part of their long-term career goals butwere willing to move for small changes in the wage rate Hence we hypothesize that high turnover rates in India aredue to rapid growth11 According to the Outsourcing Institute (wwwoutsourcingcom) these are high-growth areas in the United States

Indian expertise (outside the long-establishedbanking sector) is outdated Other skills are alsoin short supply According to the Outsourcing Insti-tute (wwwoutsourcingcom) horizontal (servicearea) expertise is most needed for BP offshoringin payroll customer care document managementand benets management Apart from account-ing skills India has little experience in thesesegments

The number size and diversity of organizationsoffshoring BPs is great The two important dimen-sions for categorizing these rms are whether therm is Indian-owned and -operated or a multina-tional or whether it is a captive or rm that under-takes outsourced work Because the potentialmarket is so large and the economics so compel-ling there have been a plethora of entrants from alarge variety of backgrounds (see Table 2) BecauseBP offshoring to India is only in its earliest stages itis difcult to predict which organizational forms willbecome dominant Moreover it is not clear whetherthere will be a single BP offshoring industry andwhether the captives or independents will dominateor even compete There are niche areas such asmedical transcription Geographic Information Sys-tems (GIS) data entry and document conversion thatmay remain separate from the industryrsquos main-stream The following general overview is not ex-haustive but does examine the types of greatestsignicance

As was the case in IT MNC captives led the wayin establishing the rst BP offshoring operations inIndia They are still considered the market leadersbecause of the in-house capabilities that they havedeveloped and the sophistication of work under-taken However in terms of employment IToutsourcing where domestic rms soon becamedominant in BP offshoring it is less clear whetherthe Indian industry domestic rms will be dominantIn chronological order the MNCs entered the mar-ket rst followed by the Indian specialists that en-tered in the late 1990s It is far more difcult toestablish the entrance dates for the myriad otherrms Finally the most recent entrants are MNCoutsourcers such as Accenture Convergys andSykes

The MNC captives include the oldest BP operationsin India and their number has increased rapidly Theearliest BP operations were established by MNCshaving existing operations in India The rst back-ofce BP operation in India was established byAmerican Express in 1993 In 1996 British Airwaysestablished a back-ofce operation in India This wasfollowed in 1998 by General Electric Today GeneralElectric is the largest BP employer with more than12000 persons in its operations and intends to in-crease to 20000 by the middle of 2004 Since2000 an increasing number of Global Fortune 500rms including AOL Citigroup Dell HewlettPackard HSBC and JP Morgan Chase have estab-lished operations

Even as MNCs with existing Indian operationsdraw on them to establish BP units MNCs that pre-viously had no Indian operations are establishing BPoperations These include a major US PC rm and alarge online Internet service provider (ISP) The new-comers are also rapidly expanding their operationsFor example in June 2001 a PC maker launched itsIndian call center operations with 200 employeesBy April 2003 it had grown to a total of 3200 em-ployees in the original facility and at a newly openedfacility in another Indian city The operations in-cluded not only a call center but also other activitiesincluding software development The growth of theISP was similarly dramatic Operations commencedin July 2002 and already by April 2003 had grownto 1500 employees and were expected to grow to2000 by the end of July 2003 The compelling sav-ings are difcult for the newcomers to resist despiteinitial difculties in managing rapid growth in a newoperating environment

Because the captives are internal operations theyhave signicant advantages First and foremost be-cause they are internal operations they haveguaranteed markets for their services that is theyhave the advantages of hierarchy Their businessvolume is a hierarchical decision and the informationdriving the decision is excellent In the case of lower-value-added routinized work the advantages ofcaptives may not be great because the decision maybe almost solely on pricemdashrisk may be minimal Inthe case of higher-value-added processes manage-ment may choose to retain them in a captive opera-tion Not surprising the initial BPs transferred were

28 Information Technologies and International Development

Lift and Shift

at the low end of the value addition spectrumHowever this need not be the end state for theIndian operation As both the Indian operationlearns and the rm becomes more comfortable withits Indian operations higher-value-added activitiesalmost certainly will be transferred For exampleGeneral Electricrsquos Indian operation has moved up thevalue-added chain and added employees doing ac-tuarial support data modeling and portfolio riskmanagement On its health care insurance opera-tions it employs 40 medical doctors to evaluate andclassify medical claims This suggests that largerMNCs will ultimately prefer to undertake BPs in cap-tive units especially for work that requires interac-tion among global employees However call centerwork which tends to be a self-sufcient processwith limited interaction among global employeescould well continue to be outsourced

Relocating back-ofce activities to India is compli-cated technically however the organizational issuescan be even more problematic Consider the unitthat is surrendering the process though it is underintense pressure to cut costs There is at a minimuma perception of increased risk as it becomes depend-ent on an Indian counterpart that is not under its di-

rect supervision This unease may be heightened formission-critical activities Also the contributing unitoften must cope with redundancies or a managerialperception that it is losing power in the wider orga-nization The recipient unit must alleviate these con-cerns during the transfer process

Operating a captive in the Indian environmentrequires signicant managerial talent For those withestablished Indian operations this is available inter-nally The new MNC entrants might experiencesignicant learning costs One difculty they face iswhether to staff the operation with expatriate exec-utives or to hire Indians Nearly all of the new en-trants choose to send some expatriates despite theexpense For these rms the expense of maintainingan expatriate may become an issue however atpresent the savings are sufciently large so as tooffset the expense

Another type of rationalization can be a part ofcreating a multinational center of excellence Fre-quently an MNCrsquos various BP operations are nation-ally based and were developed in different historicaleras thus there are varying practices for identicalfunctions Enforcing standard operating practices indifferent national environments can be difcult be-

Volume 1 Number 2 Winter 2003 29

Dossani Kenney

Table 2 Firms Involved in Business Process Activity Offshoring to India

MNC captives (Indiaexperience)

GE AMEX CitiHSBC HP

MNC N ML

MNC captives(no experience)

Dell AOL MNC N M

MNC outsourcers Sykes SitelConvergys

MNC Y SM

MNC specialistsmedical transcrip-tion GIS

HeartlandTechdata

Kampsax India

MNC Y S

Indianindependents

Daksh InfowavzEpicentre

Indian Y VSSM

Indian specialists Kale Thomson Indian Y S

Indian captive(IT industry)

Progeon WiproSpectramind HCL

BP outsourcing

Indian Y SM

Indian captive(non-IT industry)

Zenta JindalTransworld

Indian Y S

Note Large (L) 5000 medium (M) 2000ndash5000 small (S) 750ndash2000 very small (VS) 750MNC multinational corporation GE General Electric AMEX American Express Citi CitigroupHP Hewlett Packard GIS Geographic Information Systems

cause there is a constant tendency to ldquogo nativerdquoThis drift is endemic in even the best rms and maybe most pronounced in the less intensively managedparts of the national unitrsquos operations such as theback ofces The transfer of these processes to aspecialist organization dedicated to managing themnot only creates economies of scope and expertisebut also provides an opportunity for standardizationand the removal of the process from the nationaldrift Although this relocation may be resisted at thenational level for global headquarters this may beseen as a way of improving monitoring

There is naturally concomitant risk with thecentralization of particular process practices atone global center The most signicant of these isthat the global operation will lose touch with thenational environment Here the quality of communi-cation and trust with the national operation iscritical

The nal advantage is that in the future the cap-tives could offer their services on the open marketas merchant service providers This would transformthe captive from a cost center into a prot centerAlready a few of the largest captive BP operationsare considering offering services to external custom-ers This could become signicant in the futurewhen the number of activities being transferredfrom the parent rm decreases In 2003 the num-ber of internal activities available for transfer is sogreat that securing outside customers may not yetbe justied The opportunity to exploit the capabili-ties being built in India as a prot center may be-come signicant in the future

In 2003 the captives were the largest sector ofthe BP industry in India There is every reason to ex-pect this will continue for the foreseeable futureThe advantages of retaining a captive are signicantin terms of reducing risk and possible knowledgeleakage capturing prots internally and using inter-nal operations to benchmark outsourcing contractsBecause less than 10 of the Global Fortune 1000rms currently operate in India it seems likely thatmore rms will relocate activities and that the exist-ing operations will continue to expand

BP outsourcing has a long history and has grownrapidly during the last decade Estimates of the totalsize of the BP outsourcing market vary widely Forexample different consulting rms have predicted

the global market to grow to $140 billion by 2008$544 billion by 2004 and even $12 trillion by 2006(Deloitte Research 2003) The lack of consistency isremarkable The divergence in estimates is perhapsbecause denitions differ and because business ser-vice outsourcers are a polyglot category includingdata systems outsourcers such as EDS and IBM pay-roll and accounting processors such as ADP callcenter and customer relationship managers such asConvergys Sitel and Sykes large consultancy rmssuch as Accenture and many others

For the outsourcing rms globalization is notnew when the entire business space is consideredEven before the emergence of India as an offshorelocation these outsourcers had been opening ser-vice production facilities offshore in the CaribbeanLatin America and particularly Canada Beginningin about 2000 some began operations in thePhilippines

The international outsourcers established theirIndian operations in 2001 or later as a response tocompetition from the MNC captives and the Indianindependents However the MNC outsourcers havelong-established customers and enormous domainknowledge making them formidable entrants in In-dia from where they can service their existing clientsusing low-cost Indian labor These capabilities andexisting customers have permitted them to scale uptheir Indian operations extremely rapidly For exam-ple in late 2001 a leading MNC opened its rst In-dian operation in New Delhi By April 2003 thisfacility had more than 3000 employees and it wasbuilding a second facility in Bangalore that wasslated to grow to 3000 employees The expansionof their Indian operations is not constrained by alack of customers though the further growth oftheir Indian operations may lead to the scaling backor closure of North American units EstablishingIndian operations provides for substantial cost sav-ings and serves as protection from incursion intotheir customer base by the Indian rms

The ability to transfer customers to their Indianoperations while providing backup in the UnitedStates and other locations allows service-level guar-antees that rms operating only in India cannot pro-vide The conundrum for the MNC outsourcers maybe how long their customers will support higher-cost US facilities Already there have been closuresand layoffs In the short run it may be cost effective

30 Information Technologies and International Development

Lift and Shift

to continue operation of their US facilities but un-less the facilities can be transitioned to activities thatrequire spatial proximity their future may be indoubt For example in February 2003 Sykes (2003)announced the closure of its facilities in the UnitedStates and in Europe eliminating 1800 excess seatswhile its Indian subsidiary was expected to grow to1200 seats by the end of 2003

The MNC outsourcers have signicant strengthsthe most important of which is their close customerrelationships Having multiple locations provides theredundancy that some customers require Howeverbeing headquartered in the US (or in some devel-oped country) their overhead may be higher thanthat of their Indian competitors Also managing inthe Indian environment may prove difcult What iscertain is that MNC outsourcers will have to transfermore operations offshore to remain costcompetitive

India is also attracting smaller MNCs that performlabor-intensive specialty services These services arewide ranging but are based on specialized domainexpertise Though many of these are not really BPsthey are included under the broader category ofITES Examples of this type of work include medicaltranscription map digitization cartoon animationdocument entry and conversion and other labor-intensive tasks In general these businesses are in-volved in digitizing analog materials or convertinginformation from one format or media to anotherfor example taking aerial photographs and enteringthem into a mapping program Their sheer diversityis remarkable

Taken individually these activities have limitedemployment potential however in aggregate theymay be of much greater signicance For examplethere are approximately 270000 medical tran-scriptionists scattered around the United States Smalllocal rms or even individuals undertake most tran-scription Recently there has been an effort to ratio-nalize transcription however there are few technicaleconomies of scale so this has advanced haltinglyThe rationalization process might be facilitated if theactivity could be done offshore at lower cost Onedifculty is that not only are the transcriptionists de-centralized but so is the marketmdashmaking sales andmarketing difcult Thus it is not clear whether asignicant portion of the total medical transcription

ultimately will be transferred offshore despite thepossible savings What is important is to understandthat these niches offer opportunities

Another labor-intensive activity that is being relo-cated to India is map digitization Today rms andgovernments around the world are digitizing mapsthat were previously on paper Digital maps are su-perior because they are easier to update maintainand analyze The problem is that convertingthese paper maps must be done by hand a time-consuming process Moreover conversion requiresboth attention to detail and skill without low-costlabor it would be difcult for most organizations toafford digitization Given the volume of maps to bedigitized and in the future the necessity of updatingthem this could be a signicant niche

The variety of niches within which businessescould be built is remarkable Still other possibilitiesinclude legal research using Lexis-Nexis drawing oftables and gures drawing or digitizing blueprintsand so on Transcription paper-based documentdigitization database-centric research and manymore activities exist in the pores of many US organi-zations and the economy as a whole One drawbackis that in terms of the total market size many nichesmay be too small to justify transfer to India Andyet the cost pressures on these specialty rms areencouraging examination of the feasibility ofoffshoring their more mundane work There arenumerous other niches that might be attracted tothe low-cost Indian environment

The MNC specialists are a fascinating groupbecause of their sheer diversity and the likelihoodthat their decisions will be largely unnoticed becauseof each nichersquos relative insignicance Howevertheir aggregate importance could be great becauseof the sheer number of niches that exist If thesemyriad entities begin transferring activities and pro-cesses overseas in total it could have an importantimpact

Indian specialty rms are also entering elds such asmedical transcription map digitization and manu-script preparation The difculty the Indian entrantsencounter is their relative lack of domain knowl-edge For Indian rms with deep enough domainexpertise to create offerings it may be possible forthe Indian rm to transform its sales propositionfrom offering simple labor-cost arbitrage to provid-

Volume 1 Number 2 Winter 2003 31

Dossani Kenney

ing signicant value addition For example a pub-lishing rm that initially only prepared drawings forchemistry texts now offers a full range of back-ofce services including copyediting formattingand technical support The enhanced capabilityallows not only the addition of greater value butalso provides greater bargaining capacity

An example of such capability development is theMumbai rm Kale Consultants which specializes inproviding services to the airline industry OriginallyKale offered specialized airline software packageshowever in 2000 it extended its offering to includeBPs Coupling of specialized proprietary softwaretools with BP outsourcing operations meant it couldoffer a more comprehensive package For custom-ers this created an incentive to use Kale and in themeantime created a more permanent or sticky rela-tionship (personal interview 2003)

Developing domain expertise and becoming aspecialist are difcult and they have risks becausethe rm becomes dependent on a single industry oractivity And yet they also offer the potential to oc-cupy niches that may not be drawn into the extremelyferocious competition in highly commoditized sectors

A large number of Indian-owned and operated rmshave been established for the sole purpose of offer-ing BP outsourcing services to foreign rms A num-ber of these are venture capital-supported and wereformed during the Internet boom with the objectiveof providing back ofce services to US Internetrms Not surprising the collapse of the dot-comsforced these Indian service rms to rethink theirstrategies Because these rms were supplying back-ofce work such as answering e-mails and Web-related questions it was not difcult to switch theirservice offerings toward the much larger voice sec-tor Still other independents were funded by venturecapitalists as part of the enormous hype and excite-ment about BP outsourcing to India

These independents are often dependent on afew larger customers making them vulnerable tothe termination of a contract An important exampleof this is EXL Service which was the largest Indianindependent BP outsourcing rm until October2002 At that time 90 of its revenue (S Arora2002) and more than 800 of its seats were dedi-cated to the US insurance rm Conseco However

in December 2002 Conseco led for bankruptcy andthe employees dedicated to Conseco dwindled to175 by April 2003 (Verma 2003) Because of thisEXL Servicersquos growth stalled while other rmssurged by mid-2003 however EXL Service had re-covered and was once again growing

The strategic difculties are signicant Becauseof the ferocious competition and the felt necessityto expand they are under pressure to pursue manybusiness prospects However this mitigates againsttheir expressed desire to develop domain expertiseFor example the call-center-oriented rms must de-cide whether they want to specialize in in-bound orout-bound callingmdashtwo different skill sets Anotherdifculty is that the US market is the largest in theworld but sizing a facility for the US market meansthat during the day in India the facility is often idleThe independents have been able to secure somebusiness from Europe (especially England) that al-lows them to extend facility utilization however itis still difcult to use the entire facility for more than15 shifts One method for securing greater capacityutilization is to secure activities that do not requirereal-time processing12

The ultimate fate of the independents is difcultto predict and for the smaller independentsrsquo survivalwill be precarious Their greatest difculty is in mar-keting their services to foreign rms The larger in-dependents should be able to strengthen their USmarketing thus increasing their market share Themiddle-tier independents might be acquired eitherby Indian rms or multinationals wishing to enterquickly the BP outsourcing eld The strongest inde-pendents may be able to create rms that resemblethe multinational independents

The Indian IT industry has grown remarkably rapidlyover the last decade through the provision ofoutsourced programming and IT services to theglobal market (A Arora and Athreye 2002 DrsquoCosta2003) In IT outsourcing Indian rms such as HCLInfosys Satyam TCS and Wipro have become glob-ally competitive Because of their ability to uselower-cost Indian software talent they have madesignicant global market share gains Furthermoretheir interaction with the global economy has con-tributed to the development of executive and mana-gerial talent capable of securing overseas contracts

32 Information Technologies and International Development

Lift and Shift

12 This section is based on interviews with Indian executives

managing the interface with foreign customers andmigrating activities across national and rm bound-aries In the process these rms have cultivatedclose connections with foreign customers for IT ser-vices This provides an entree and condence on thepart of customers that facilitates persuading foreigncustomers to trust the outsourcing rms with otherservices

Given the growth in ITES the Indian IT rms be-lieve BP outsourcing is a sector in which they can ex-pand Their strategic question has been how toenter this new industry The major rms have an-swered this question differently Infosys and Satyamestablished subsidiaries one of which Progeon hasgrown rapidly and recently split a 5-year $160 mil-lion contract from British Telecom with HCL BPoutsourcing In contrast the Satyam subsidiary hasexperienced only limited growth TCS the largestIndian software rm entered the BP outsourcingsector through a joint venture Finally Wipro andHCL entered the industry through acquisitionsWipro acquired a venture-funded Indian independ-ent Spectramind which has grown quickly HCLacquired the Northern Ireland call center subsidiaryof British Telecom though most of HCLrsquos BP out-sourcing growth has been in India13

The Indian IT rms have signicant advantages interms of an ability to invest linkages to customersand various other strengths However the BP out-sourcing business is different from IT In terms ofmarketing the customerrsquos key decision maker is notthe chief information ofcer or chief technicalofcer Usually BP outsourcing is sold to the variousresponsible divisions or departments Furthermorethe ultimate decision rests with the chief financialofcer or chief executive ofcer This means thatthe Indian rm must operate through differentchannels

The BP workforce is also different Whereas inthe IT sector the workforce largely consists of engi-neers in BPs the workersrsquo degrees are in commerceand social science Because BP outsourcing work of-ten requires direct interaction with customers theworkforce-salient skills are interpersonal rather thantechnical Furthermore many BPs are undertaken inreal time therefore errors and mistakes have an im-mediate impact Service-level agreements are tightlywritten and monitored in real time therefore prob-lems are exposed nearly immediately In software

mistakes can be rectied later Also BPs that requirecustomer interaction can be extremely stressfulputting a premium on skillful workforcemanagement

The ability of Indian IT rms to manage nontech-nical personnel in extremely price-competitive envi-ronments will be tested However any test of theirmanagerial prowess may not come until later as therapid market growth will initially ensure an appear-ance of success for most entrants Difculties mayremain hidden until growth slows though by thattime they may have built such close relationshipswith their customers that exit by the customers mayno longer be possible because of its depth and stick-iness There is also the possibility that the technicalskills within the IT parent could be used to automateaspects of the BP outsourcing process creating an-other level of value addition that would improveprotability This would also enable the IT rm sub-sidiaries to create advantages beyond routine labor-cost arbitrage

There is a host of other established Indian rms en-tering the BP outsourcing industry Attracted by theldquoGold Rushrdquo aspects of the sector these traditionalrms with their roots in the large Indian businessgroups have invested signicant sums Already someof them appear to be experiencing difculties in se-curing customers (personal interviews 2003) How-ever in contrast with the smaller independentsthese captives have deep pockets and can competefor as long as their parents are willing to providesubsidies They will either nd a successful strategyor they will exit the business because of an unwill-ingness of the parent to sustain further losses

This genre of captives is interesting because oftenthey have no particular advantages In almost allcases there are few synergies between the parentrsquosexisting businesses and the services they aim to pro-vide They nearly always have experienced manage-ment though their experience may be in therelatively protected domestic market Frequentlythey have minimal experience in interacting withforeign clients especially in terms of providing ser-vices The lack of inherent advantages beyond deeppockets means that these captives will have to buildcapabilities in the same way as the Indian independ-ents Their only signicant advantage will be the rel-

Volume 1 Number 2 Winter 2003 33

Dossani Kenney

13 In June 2003 BT announced that it was establishing a captive Indian call center

ative deep pockets of their parent rms thoughoddly enough this may inhibit their ability to evolveto market demands In other words protection fromthe vagaries of the market could contribute to aninability to learn from the market

During the next decade the big debate about glob-alization will no longer concentrate on manufactur-ing which is an increasingly small portion of ourlabor force14 The globalization of services will takemanufacturingrsquos place The constantly falling price ofbandwidth the increasing digitization of informa-tion and the drive to reengineer and modularizeservices are making it possible to lift and shift themoverseas We argue that the favored destination isIndia and that at this time the rate of growth inthis transfer is signicant

The movement of business processes offshore isnot new Until the 1990s the undertaking of busi-ness processes in developing nations existed but wasminimal This paper shows how this may be in theprocess of changing dramatically We have beencareful not to exaggerate the current state of BPoffshoring Most rms have been cautious in theirdecisions to transfer these activities (especially mis-sion-critical functions) offshore especially to nationssuch as India or the Philippines that have differentsocial and political arrangements and sometimesproblematic physical infrastructures There are alsointernal risks that come from separating various ac-tivities in a process and possible resistance from de-veloped country workforces Finally a botchedtransfer process can be traumatic to the rmrsquos cus-tomers and employees

If there are dangers in offshoring BPs there isalso the potential for enormous savings In every in-terview and through our calculations there is nodoubt that a well-handled transfer of an activity cangenerate at least a 40 savings on the entire pro-cess (not just on labor costs) Table 3 presents thecost differences between operating a call center inKansas City Missouri in the United States and a callcenter in Bangalore India One rm claimed thattheir savings over a 2-year period were as high as80 for some processes The evidence we werepermitted to see but not reproduce suggests

strongly that quality is as good as or better thanthat achieved in comparable US facilities

These savings and quality claims cannot be inde-pendently veried however during the last 3 yearsthe overall growth rate in employment of BP opera-tions has been more than 50 per annum More-over every MNC we interviewed was dramaticallyincreasing the size of its Indian operations andmany said that the bottleneck was their ability togrow even more quickly The larger Indian independ-ents and Indian IT industry captives were growing ata similar pace Finally many other newcomers werebuilding facilities or exploring India with the aim ofestablishing operations This is prima facie evidencethat there are signicant economic benets Havingsaid that we are careful not to assume that this ratewill continue indenitely We identied several fac-tors that must be considered before concluding thata particular business process can be offshored in-cluding the complexity of the task the level ofinteractivity its knowledge component and level ofseparability the possibility of reengineering the ad-vantages of a single location and of scale and thetime-sensitive nature of the task However thereseems to be an enormous opportunity for furthergrowth when one considers that in 2002 more than85 million Americans were in the service sector (USBureau of Labor Statistics 2003) If only 2 of thoseservice jobs could be transferred this would be 17million jobs and the savings would be enormousOur opinion is that 2 is not an unreasonableestimate

In this study we identied seven types of BP op-erations in India There is no reason to believe thatonly one type can survive At the present the MNCcaptives have the greatest number of employeesundertake the highest-value-added work and areable to get the greatest usage of each of their seatsControlling the process internally mitigates the risksthat can come from outsourcing Finally it allowsthe MNC to capture all the prots of offshoringMost MNC outsourcers should also succeed in Indiaas they will be able to provide their developed na-tion customers with security and with their globalpresence a guarantee of redundancy Their greatestdanger is that BP offshoring to India becomes soroutinized that they will have to compete with thelower-cost Indian outsourcers directly In general the

34 Information Technologies and International Development

Lift and Shift

14 For a recent discussion of globalization in various manufacturing industries see Kenney and Florida (2004)

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

cause there is a constant tendency to ldquogo nativerdquoThis drift is endemic in even the best rms and maybe most pronounced in the less intensively managedparts of the national unitrsquos operations such as theback ofces The transfer of these processes to aspecialist organization dedicated to managing themnot only creates economies of scope and expertisebut also provides an opportunity for standardizationand the removal of the process from the nationaldrift Although this relocation may be resisted at thenational level for global headquarters this may beseen as a way of improving monitoring

There is naturally concomitant risk with thecentralization of particular process practices atone global center The most signicant of these isthat the global operation will lose touch with thenational environment Here the quality of communi-cation and trust with the national operation iscritical

The nal advantage is that in the future the cap-tives could offer their services on the open marketas merchant service providers This would transformthe captive from a cost center into a prot centerAlready a few of the largest captive BP operationsare considering offering services to external custom-ers This could become signicant in the futurewhen the number of activities being transferredfrom the parent rm decreases In 2003 the num-ber of internal activities available for transfer is sogreat that securing outside customers may not yetbe justied The opportunity to exploit the capabili-ties being built in India as a prot center may be-come signicant in the future

In 2003 the captives were the largest sector ofthe BP industry in India There is every reason to ex-pect this will continue for the foreseeable futureThe advantages of retaining a captive are signicantin terms of reducing risk and possible knowledgeleakage capturing prots internally and using inter-nal operations to benchmark outsourcing contractsBecause less than 10 of the Global Fortune 1000rms currently operate in India it seems likely thatmore rms will relocate activities and that the exist-ing operations will continue to expand

BP outsourcing has a long history and has grownrapidly during the last decade Estimates of the totalsize of the BP outsourcing market vary widely Forexample different consulting rms have predicted

the global market to grow to $140 billion by 2008$544 billion by 2004 and even $12 trillion by 2006(Deloitte Research 2003) The lack of consistency isremarkable The divergence in estimates is perhapsbecause denitions differ and because business ser-vice outsourcers are a polyglot category includingdata systems outsourcers such as EDS and IBM pay-roll and accounting processors such as ADP callcenter and customer relationship managers such asConvergys Sitel and Sykes large consultancy rmssuch as Accenture and many others

For the outsourcing rms globalization is notnew when the entire business space is consideredEven before the emergence of India as an offshorelocation these outsourcers had been opening ser-vice production facilities offshore in the CaribbeanLatin America and particularly Canada Beginningin about 2000 some began operations in thePhilippines

The international outsourcers established theirIndian operations in 2001 or later as a response tocompetition from the MNC captives and the Indianindependents However the MNC outsourcers havelong-established customers and enormous domainknowledge making them formidable entrants in In-dia from where they can service their existing clientsusing low-cost Indian labor These capabilities andexisting customers have permitted them to scale uptheir Indian operations extremely rapidly For exam-ple in late 2001 a leading MNC opened its rst In-dian operation in New Delhi By April 2003 thisfacility had more than 3000 employees and it wasbuilding a second facility in Bangalore that wasslated to grow to 3000 employees The expansionof their Indian operations is not constrained by alack of customers though the further growth oftheir Indian operations may lead to the scaling backor closure of North American units EstablishingIndian operations provides for substantial cost sav-ings and serves as protection from incursion intotheir customer base by the Indian rms

The ability to transfer customers to their Indianoperations while providing backup in the UnitedStates and other locations allows service-level guar-antees that rms operating only in India cannot pro-vide The conundrum for the MNC outsourcers maybe how long their customers will support higher-cost US facilities Already there have been closuresand layoffs In the short run it may be cost effective

30 Information Technologies and International Development

Lift and Shift

to continue operation of their US facilities but un-less the facilities can be transitioned to activities thatrequire spatial proximity their future may be indoubt For example in February 2003 Sykes (2003)announced the closure of its facilities in the UnitedStates and in Europe eliminating 1800 excess seatswhile its Indian subsidiary was expected to grow to1200 seats by the end of 2003

The MNC outsourcers have signicant strengthsthe most important of which is their close customerrelationships Having multiple locations provides theredundancy that some customers require Howeverbeing headquartered in the US (or in some devel-oped country) their overhead may be higher thanthat of their Indian competitors Also managing inthe Indian environment may prove difcult What iscertain is that MNC outsourcers will have to transfermore operations offshore to remain costcompetitive

India is also attracting smaller MNCs that performlabor-intensive specialty services These services arewide ranging but are based on specialized domainexpertise Though many of these are not really BPsthey are included under the broader category ofITES Examples of this type of work include medicaltranscription map digitization cartoon animationdocument entry and conversion and other labor-intensive tasks In general these businesses are in-volved in digitizing analog materials or convertinginformation from one format or media to anotherfor example taking aerial photographs and enteringthem into a mapping program Their sheer diversityis remarkable

Taken individually these activities have limitedemployment potential however in aggregate theymay be of much greater signicance For examplethere are approximately 270000 medical tran-scriptionists scattered around the United States Smalllocal rms or even individuals undertake most tran-scription Recently there has been an effort to ratio-nalize transcription however there are few technicaleconomies of scale so this has advanced haltinglyThe rationalization process might be facilitated if theactivity could be done offshore at lower cost Onedifculty is that not only are the transcriptionists de-centralized but so is the marketmdashmaking sales andmarketing difcult Thus it is not clear whether asignicant portion of the total medical transcription

ultimately will be transferred offshore despite thepossible savings What is important is to understandthat these niches offer opportunities

Another labor-intensive activity that is being relo-cated to India is map digitization Today rms andgovernments around the world are digitizing mapsthat were previously on paper Digital maps are su-perior because they are easier to update maintainand analyze The problem is that convertingthese paper maps must be done by hand a time-consuming process Moreover conversion requiresboth attention to detail and skill without low-costlabor it would be difcult for most organizations toafford digitization Given the volume of maps to bedigitized and in the future the necessity of updatingthem this could be a signicant niche

The variety of niches within which businessescould be built is remarkable Still other possibilitiesinclude legal research using Lexis-Nexis drawing oftables and gures drawing or digitizing blueprintsand so on Transcription paper-based documentdigitization database-centric research and manymore activities exist in the pores of many US organi-zations and the economy as a whole One drawbackis that in terms of the total market size many nichesmay be too small to justify transfer to India Andyet the cost pressures on these specialty rms areencouraging examination of the feasibility ofoffshoring their more mundane work There arenumerous other niches that might be attracted tothe low-cost Indian environment

The MNC specialists are a fascinating groupbecause of their sheer diversity and the likelihoodthat their decisions will be largely unnoticed becauseof each nichersquos relative insignicance Howevertheir aggregate importance could be great becauseof the sheer number of niches that exist If thesemyriad entities begin transferring activities and pro-cesses overseas in total it could have an importantimpact

Indian specialty rms are also entering elds such asmedical transcription map digitization and manu-script preparation The difculty the Indian entrantsencounter is their relative lack of domain knowl-edge For Indian rms with deep enough domainexpertise to create offerings it may be possible forthe Indian rm to transform its sales propositionfrom offering simple labor-cost arbitrage to provid-

Volume 1 Number 2 Winter 2003 31

Dossani Kenney

ing signicant value addition For example a pub-lishing rm that initially only prepared drawings forchemistry texts now offers a full range of back-ofce services including copyediting formattingand technical support The enhanced capabilityallows not only the addition of greater value butalso provides greater bargaining capacity

An example of such capability development is theMumbai rm Kale Consultants which specializes inproviding services to the airline industry OriginallyKale offered specialized airline software packageshowever in 2000 it extended its offering to includeBPs Coupling of specialized proprietary softwaretools with BP outsourcing operations meant it couldoffer a more comprehensive package For custom-ers this created an incentive to use Kale and in themeantime created a more permanent or sticky rela-tionship (personal interview 2003)

Developing domain expertise and becoming aspecialist are difcult and they have risks becausethe rm becomes dependent on a single industry oractivity And yet they also offer the potential to oc-cupy niches that may not be drawn into the extremelyferocious competition in highly commoditized sectors

A large number of Indian-owned and operated rmshave been established for the sole purpose of offer-ing BP outsourcing services to foreign rms A num-ber of these are venture capital-supported and wereformed during the Internet boom with the objectiveof providing back ofce services to US Internetrms Not surprising the collapse of the dot-comsforced these Indian service rms to rethink theirstrategies Because these rms were supplying back-ofce work such as answering e-mails and Web-related questions it was not difcult to switch theirservice offerings toward the much larger voice sec-tor Still other independents were funded by venturecapitalists as part of the enormous hype and excite-ment about BP outsourcing to India

These independents are often dependent on afew larger customers making them vulnerable tothe termination of a contract An important exampleof this is EXL Service which was the largest Indianindependent BP outsourcing rm until October2002 At that time 90 of its revenue (S Arora2002) and more than 800 of its seats were dedi-cated to the US insurance rm Conseco However

in December 2002 Conseco led for bankruptcy andthe employees dedicated to Conseco dwindled to175 by April 2003 (Verma 2003) Because of thisEXL Servicersquos growth stalled while other rmssurged by mid-2003 however EXL Service had re-covered and was once again growing

The strategic difculties are signicant Becauseof the ferocious competition and the felt necessityto expand they are under pressure to pursue manybusiness prospects However this mitigates againsttheir expressed desire to develop domain expertiseFor example the call-center-oriented rms must de-cide whether they want to specialize in in-bound orout-bound callingmdashtwo different skill sets Anotherdifculty is that the US market is the largest in theworld but sizing a facility for the US market meansthat during the day in India the facility is often idleThe independents have been able to secure somebusiness from Europe (especially England) that al-lows them to extend facility utilization however itis still difcult to use the entire facility for more than15 shifts One method for securing greater capacityutilization is to secure activities that do not requirereal-time processing12

The ultimate fate of the independents is difcultto predict and for the smaller independentsrsquo survivalwill be precarious Their greatest difculty is in mar-keting their services to foreign rms The larger in-dependents should be able to strengthen their USmarketing thus increasing their market share Themiddle-tier independents might be acquired eitherby Indian rms or multinationals wishing to enterquickly the BP outsourcing eld The strongest inde-pendents may be able to create rms that resemblethe multinational independents

The Indian IT industry has grown remarkably rapidlyover the last decade through the provision ofoutsourced programming and IT services to theglobal market (A Arora and Athreye 2002 DrsquoCosta2003) In IT outsourcing Indian rms such as HCLInfosys Satyam TCS and Wipro have become glob-ally competitive Because of their ability to uselower-cost Indian software talent they have madesignicant global market share gains Furthermoretheir interaction with the global economy has con-tributed to the development of executive and mana-gerial talent capable of securing overseas contracts

32 Information Technologies and International Development

Lift and Shift

12 This section is based on interviews with Indian executives

managing the interface with foreign customers andmigrating activities across national and rm bound-aries In the process these rms have cultivatedclose connections with foreign customers for IT ser-vices This provides an entree and condence on thepart of customers that facilitates persuading foreigncustomers to trust the outsourcing rms with otherservices

Given the growth in ITES the Indian IT rms be-lieve BP outsourcing is a sector in which they can ex-pand Their strategic question has been how toenter this new industry The major rms have an-swered this question differently Infosys and Satyamestablished subsidiaries one of which Progeon hasgrown rapidly and recently split a 5-year $160 mil-lion contract from British Telecom with HCL BPoutsourcing In contrast the Satyam subsidiary hasexperienced only limited growth TCS the largestIndian software rm entered the BP outsourcingsector through a joint venture Finally Wipro andHCL entered the industry through acquisitionsWipro acquired a venture-funded Indian independ-ent Spectramind which has grown quickly HCLacquired the Northern Ireland call center subsidiaryof British Telecom though most of HCLrsquos BP out-sourcing growth has been in India13

The Indian IT rms have signicant advantages interms of an ability to invest linkages to customersand various other strengths However the BP out-sourcing business is different from IT In terms ofmarketing the customerrsquos key decision maker is notthe chief information ofcer or chief technicalofcer Usually BP outsourcing is sold to the variousresponsible divisions or departments Furthermorethe ultimate decision rests with the chief financialofcer or chief executive ofcer This means thatthe Indian rm must operate through differentchannels

The BP workforce is also different Whereas inthe IT sector the workforce largely consists of engi-neers in BPs the workersrsquo degrees are in commerceand social science Because BP outsourcing work of-ten requires direct interaction with customers theworkforce-salient skills are interpersonal rather thantechnical Furthermore many BPs are undertaken inreal time therefore errors and mistakes have an im-mediate impact Service-level agreements are tightlywritten and monitored in real time therefore prob-lems are exposed nearly immediately In software

mistakes can be rectied later Also BPs that requirecustomer interaction can be extremely stressfulputting a premium on skillful workforcemanagement

The ability of Indian IT rms to manage nontech-nical personnel in extremely price-competitive envi-ronments will be tested However any test of theirmanagerial prowess may not come until later as therapid market growth will initially ensure an appear-ance of success for most entrants Difculties mayremain hidden until growth slows though by thattime they may have built such close relationshipswith their customers that exit by the customers mayno longer be possible because of its depth and stick-iness There is also the possibility that the technicalskills within the IT parent could be used to automateaspects of the BP outsourcing process creating an-other level of value addition that would improveprotability This would also enable the IT rm sub-sidiaries to create advantages beyond routine labor-cost arbitrage

There is a host of other established Indian rms en-tering the BP outsourcing industry Attracted by theldquoGold Rushrdquo aspects of the sector these traditionalrms with their roots in the large Indian businessgroups have invested signicant sums Already someof them appear to be experiencing difculties in se-curing customers (personal interviews 2003) How-ever in contrast with the smaller independentsthese captives have deep pockets and can competefor as long as their parents are willing to providesubsidies They will either nd a successful strategyor they will exit the business because of an unwill-ingness of the parent to sustain further losses

This genre of captives is interesting because oftenthey have no particular advantages In almost allcases there are few synergies between the parentrsquosexisting businesses and the services they aim to pro-vide They nearly always have experienced manage-ment though their experience may be in therelatively protected domestic market Frequentlythey have minimal experience in interacting withforeign clients especially in terms of providing ser-vices The lack of inherent advantages beyond deeppockets means that these captives will have to buildcapabilities in the same way as the Indian independ-ents Their only signicant advantage will be the rel-

Volume 1 Number 2 Winter 2003 33

Dossani Kenney

13 In June 2003 BT announced that it was establishing a captive Indian call center

ative deep pockets of their parent rms thoughoddly enough this may inhibit their ability to evolveto market demands In other words protection fromthe vagaries of the market could contribute to aninability to learn from the market

During the next decade the big debate about glob-alization will no longer concentrate on manufactur-ing which is an increasingly small portion of ourlabor force14 The globalization of services will takemanufacturingrsquos place The constantly falling price ofbandwidth the increasing digitization of informa-tion and the drive to reengineer and modularizeservices are making it possible to lift and shift themoverseas We argue that the favored destination isIndia and that at this time the rate of growth inthis transfer is signicant

The movement of business processes offshore isnot new Until the 1990s the undertaking of busi-ness processes in developing nations existed but wasminimal This paper shows how this may be in theprocess of changing dramatically We have beencareful not to exaggerate the current state of BPoffshoring Most rms have been cautious in theirdecisions to transfer these activities (especially mis-sion-critical functions) offshore especially to nationssuch as India or the Philippines that have differentsocial and political arrangements and sometimesproblematic physical infrastructures There are alsointernal risks that come from separating various ac-tivities in a process and possible resistance from de-veloped country workforces Finally a botchedtransfer process can be traumatic to the rmrsquos cus-tomers and employees

If there are dangers in offshoring BPs there isalso the potential for enormous savings In every in-terview and through our calculations there is nodoubt that a well-handled transfer of an activity cangenerate at least a 40 savings on the entire pro-cess (not just on labor costs) Table 3 presents thecost differences between operating a call center inKansas City Missouri in the United States and a callcenter in Bangalore India One rm claimed thattheir savings over a 2-year period were as high as80 for some processes The evidence we werepermitted to see but not reproduce suggests

strongly that quality is as good as or better thanthat achieved in comparable US facilities

These savings and quality claims cannot be inde-pendently veried however during the last 3 yearsthe overall growth rate in employment of BP opera-tions has been more than 50 per annum More-over every MNC we interviewed was dramaticallyincreasing the size of its Indian operations andmany said that the bottleneck was their ability togrow even more quickly The larger Indian independ-ents and Indian IT industry captives were growing ata similar pace Finally many other newcomers werebuilding facilities or exploring India with the aim ofestablishing operations This is prima facie evidencethat there are signicant economic benets Havingsaid that we are careful not to assume that this ratewill continue indenitely We identied several fac-tors that must be considered before concluding thata particular business process can be offshored in-cluding the complexity of the task the level ofinteractivity its knowledge component and level ofseparability the possibility of reengineering the ad-vantages of a single location and of scale and thetime-sensitive nature of the task However thereseems to be an enormous opportunity for furthergrowth when one considers that in 2002 more than85 million Americans were in the service sector (USBureau of Labor Statistics 2003) If only 2 of thoseservice jobs could be transferred this would be 17million jobs and the savings would be enormousOur opinion is that 2 is not an unreasonableestimate

In this study we identied seven types of BP op-erations in India There is no reason to believe thatonly one type can survive At the present the MNCcaptives have the greatest number of employeesundertake the highest-value-added work and areable to get the greatest usage of each of their seatsControlling the process internally mitigates the risksthat can come from outsourcing Finally it allowsthe MNC to capture all the prots of offshoringMost MNC outsourcers should also succeed in Indiaas they will be able to provide their developed na-tion customers with security and with their globalpresence a guarantee of redundancy Their greatestdanger is that BP offshoring to India becomes soroutinized that they will have to compete with thelower-cost Indian outsourcers directly In general the

34 Information Technologies and International Development

Lift and Shift

14 For a recent discussion of globalization in various manufacturing industries see Kenney and Florida (2004)

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

to continue operation of their US facilities but un-less the facilities can be transitioned to activities thatrequire spatial proximity their future may be indoubt For example in February 2003 Sykes (2003)announced the closure of its facilities in the UnitedStates and in Europe eliminating 1800 excess seatswhile its Indian subsidiary was expected to grow to1200 seats by the end of 2003

The MNC outsourcers have signicant strengthsthe most important of which is their close customerrelationships Having multiple locations provides theredundancy that some customers require Howeverbeing headquartered in the US (or in some devel-oped country) their overhead may be higher thanthat of their Indian competitors Also managing inthe Indian environment may prove difcult What iscertain is that MNC outsourcers will have to transfermore operations offshore to remain costcompetitive

India is also attracting smaller MNCs that performlabor-intensive specialty services These services arewide ranging but are based on specialized domainexpertise Though many of these are not really BPsthey are included under the broader category ofITES Examples of this type of work include medicaltranscription map digitization cartoon animationdocument entry and conversion and other labor-intensive tasks In general these businesses are in-volved in digitizing analog materials or convertinginformation from one format or media to anotherfor example taking aerial photographs and enteringthem into a mapping program Their sheer diversityis remarkable

Taken individually these activities have limitedemployment potential however in aggregate theymay be of much greater signicance For examplethere are approximately 270000 medical tran-scriptionists scattered around the United States Smalllocal rms or even individuals undertake most tran-scription Recently there has been an effort to ratio-nalize transcription however there are few technicaleconomies of scale so this has advanced haltinglyThe rationalization process might be facilitated if theactivity could be done offshore at lower cost Onedifculty is that not only are the transcriptionists de-centralized but so is the marketmdashmaking sales andmarketing difcult Thus it is not clear whether asignicant portion of the total medical transcription

ultimately will be transferred offshore despite thepossible savings What is important is to understandthat these niches offer opportunities

Another labor-intensive activity that is being relo-cated to India is map digitization Today rms andgovernments around the world are digitizing mapsthat were previously on paper Digital maps are su-perior because they are easier to update maintainand analyze The problem is that convertingthese paper maps must be done by hand a time-consuming process Moreover conversion requiresboth attention to detail and skill without low-costlabor it would be difcult for most organizations toafford digitization Given the volume of maps to bedigitized and in the future the necessity of updatingthem this could be a signicant niche

The variety of niches within which businessescould be built is remarkable Still other possibilitiesinclude legal research using Lexis-Nexis drawing oftables and gures drawing or digitizing blueprintsand so on Transcription paper-based documentdigitization database-centric research and manymore activities exist in the pores of many US organi-zations and the economy as a whole One drawbackis that in terms of the total market size many nichesmay be too small to justify transfer to India Andyet the cost pressures on these specialty rms areencouraging examination of the feasibility ofoffshoring their more mundane work There arenumerous other niches that might be attracted tothe low-cost Indian environment

The MNC specialists are a fascinating groupbecause of their sheer diversity and the likelihoodthat their decisions will be largely unnoticed becauseof each nichersquos relative insignicance Howevertheir aggregate importance could be great becauseof the sheer number of niches that exist If thesemyriad entities begin transferring activities and pro-cesses overseas in total it could have an importantimpact

Indian specialty rms are also entering elds such asmedical transcription map digitization and manu-script preparation The difculty the Indian entrantsencounter is their relative lack of domain knowl-edge For Indian rms with deep enough domainexpertise to create offerings it may be possible forthe Indian rm to transform its sales propositionfrom offering simple labor-cost arbitrage to provid-

Volume 1 Number 2 Winter 2003 31

Dossani Kenney

ing signicant value addition For example a pub-lishing rm that initially only prepared drawings forchemistry texts now offers a full range of back-ofce services including copyediting formattingand technical support The enhanced capabilityallows not only the addition of greater value butalso provides greater bargaining capacity

An example of such capability development is theMumbai rm Kale Consultants which specializes inproviding services to the airline industry OriginallyKale offered specialized airline software packageshowever in 2000 it extended its offering to includeBPs Coupling of specialized proprietary softwaretools with BP outsourcing operations meant it couldoffer a more comprehensive package For custom-ers this created an incentive to use Kale and in themeantime created a more permanent or sticky rela-tionship (personal interview 2003)

Developing domain expertise and becoming aspecialist are difcult and they have risks becausethe rm becomes dependent on a single industry oractivity And yet they also offer the potential to oc-cupy niches that may not be drawn into the extremelyferocious competition in highly commoditized sectors

A large number of Indian-owned and operated rmshave been established for the sole purpose of offer-ing BP outsourcing services to foreign rms A num-ber of these are venture capital-supported and wereformed during the Internet boom with the objectiveof providing back ofce services to US Internetrms Not surprising the collapse of the dot-comsforced these Indian service rms to rethink theirstrategies Because these rms were supplying back-ofce work such as answering e-mails and Web-related questions it was not difcult to switch theirservice offerings toward the much larger voice sec-tor Still other independents were funded by venturecapitalists as part of the enormous hype and excite-ment about BP outsourcing to India

These independents are often dependent on afew larger customers making them vulnerable tothe termination of a contract An important exampleof this is EXL Service which was the largest Indianindependent BP outsourcing rm until October2002 At that time 90 of its revenue (S Arora2002) and more than 800 of its seats were dedi-cated to the US insurance rm Conseco However

in December 2002 Conseco led for bankruptcy andthe employees dedicated to Conseco dwindled to175 by April 2003 (Verma 2003) Because of thisEXL Servicersquos growth stalled while other rmssurged by mid-2003 however EXL Service had re-covered and was once again growing

The strategic difculties are signicant Becauseof the ferocious competition and the felt necessityto expand they are under pressure to pursue manybusiness prospects However this mitigates againsttheir expressed desire to develop domain expertiseFor example the call-center-oriented rms must de-cide whether they want to specialize in in-bound orout-bound callingmdashtwo different skill sets Anotherdifculty is that the US market is the largest in theworld but sizing a facility for the US market meansthat during the day in India the facility is often idleThe independents have been able to secure somebusiness from Europe (especially England) that al-lows them to extend facility utilization however itis still difcult to use the entire facility for more than15 shifts One method for securing greater capacityutilization is to secure activities that do not requirereal-time processing12

The ultimate fate of the independents is difcultto predict and for the smaller independentsrsquo survivalwill be precarious Their greatest difculty is in mar-keting their services to foreign rms The larger in-dependents should be able to strengthen their USmarketing thus increasing their market share Themiddle-tier independents might be acquired eitherby Indian rms or multinationals wishing to enterquickly the BP outsourcing eld The strongest inde-pendents may be able to create rms that resemblethe multinational independents

The Indian IT industry has grown remarkably rapidlyover the last decade through the provision ofoutsourced programming and IT services to theglobal market (A Arora and Athreye 2002 DrsquoCosta2003) In IT outsourcing Indian rms such as HCLInfosys Satyam TCS and Wipro have become glob-ally competitive Because of their ability to uselower-cost Indian software talent they have madesignicant global market share gains Furthermoretheir interaction with the global economy has con-tributed to the development of executive and mana-gerial talent capable of securing overseas contracts

32 Information Technologies and International Development

Lift and Shift

12 This section is based on interviews with Indian executives

managing the interface with foreign customers andmigrating activities across national and rm bound-aries In the process these rms have cultivatedclose connections with foreign customers for IT ser-vices This provides an entree and condence on thepart of customers that facilitates persuading foreigncustomers to trust the outsourcing rms with otherservices

Given the growth in ITES the Indian IT rms be-lieve BP outsourcing is a sector in which they can ex-pand Their strategic question has been how toenter this new industry The major rms have an-swered this question differently Infosys and Satyamestablished subsidiaries one of which Progeon hasgrown rapidly and recently split a 5-year $160 mil-lion contract from British Telecom with HCL BPoutsourcing In contrast the Satyam subsidiary hasexperienced only limited growth TCS the largestIndian software rm entered the BP outsourcingsector through a joint venture Finally Wipro andHCL entered the industry through acquisitionsWipro acquired a venture-funded Indian independ-ent Spectramind which has grown quickly HCLacquired the Northern Ireland call center subsidiaryof British Telecom though most of HCLrsquos BP out-sourcing growth has been in India13

The Indian IT rms have signicant advantages interms of an ability to invest linkages to customersand various other strengths However the BP out-sourcing business is different from IT In terms ofmarketing the customerrsquos key decision maker is notthe chief information ofcer or chief technicalofcer Usually BP outsourcing is sold to the variousresponsible divisions or departments Furthermorethe ultimate decision rests with the chief financialofcer or chief executive ofcer This means thatthe Indian rm must operate through differentchannels

The BP workforce is also different Whereas inthe IT sector the workforce largely consists of engi-neers in BPs the workersrsquo degrees are in commerceand social science Because BP outsourcing work of-ten requires direct interaction with customers theworkforce-salient skills are interpersonal rather thantechnical Furthermore many BPs are undertaken inreal time therefore errors and mistakes have an im-mediate impact Service-level agreements are tightlywritten and monitored in real time therefore prob-lems are exposed nearly immediately In software

mistakes can be rectied later Also BPs that requirecustomer interaction can be extremely stressfulputting a premium on skillful workforcemanagement

The ability of Indian IT rms to manage nontech-nical personnel in extremely price-competitive envi-ronments will be tested However any test of theirmanagerial prowess may not come until later as therapid market growth will initially ensure an appear-ance of success for most entrants Difculties mayremain hidden until growth slows though by thattime they may have built such close relationshipswith their customers that exit by the customers mayno longer be possible because of its depth and stick-iness There is also the possibility that the technicalskills within the IT parent could be used to automateaspects of the BP outsourcing process creating an-other level of value addition that would improveprotability This would also enable the IT rm sub-sidiaries to create advantages beyond routine labor-cost arbitrage

There is a host of other established Indian rms en-tering the BP outsourcing industry Attracted by theldquoGold Rushrdquo aspects of the sector these traditionalrms with their roots in the large Indian businessgroups have invested signicant sums Already someof them appear to be experiencing difculties in se-curing customers (personal interviews 2003) How-ever in contrast with the smaller independentsthese captives have deep pockets and can competefor as long as their parents are willing to providesubsidies They will either nd a successful strategyor they will exit the business because of an unwill-ingness of the parent to sustain further losses

This genre of captives is interesting because oftenthey have no particular advantages In almost allcases there are few synergies between the parentrsquosexisting businesses and the services they aim to pro-vide They nearly always have experienced manage-ment though their experience may be in therelatively protected domestic market Frequentlythey have minimal experience in interacting withforeign clients especially in terms of providing ser-vices The lack of inherent advantages beyond deeppockets means that these captives will have to buildcapabilities in the same way as the Indian independ-ents Their only signicant advantage will be the rel-

Volume 1 Number 2 Winter 2003 33

Dossani Kenney

13 In June 2003 BT announced that it was establishing a captive Indian call center

ative deep pockets of their parent rms thoughoddly enough this may inhibit their ability to evolveto market demands In other words protection fromthe vagaries of the market could contribute to aninability to learn from the market

During the next decade the big debate about glob-alization will no longer concentrate on manufactur-ing which is an increasingly small portion of ourlabor force14 The globalization of services will takemanufacturingrsquos place The constantly falling price ofbandwidth the increasing digitization of informa-tion and the drive to reengineer and modularizeservices are making it possible to lift and shift themoverseas We argue that the favored destination isIndia and that at this time the rate of growth inthis transfer is signicant

The movement of business processes offshore isnot new Until the 1990s the undertaking of busi-ness processes in developing nations existed but wasminimal This paper shows how this may be in theprocess of changing dramatically We have beencareful not to exaggerate the current state of BPoffshoring Most rms have been cautious in theirdecisions to transfer these activities (especially mis-sion-critical functions) offshore especially to nationssuch as India or the Philippines that have differentsocial and political arrangements and sometimesproblematic physical infrastructures There are alsointernal risks that come from separating various ac-tivities in a process and possible resistance from de-veloped country workforces Finally a botchedtransfer process can be traumatic to the rmrsquos cus-tomers and employees

If there are dangers in offshoring BPs there isalso the potential for enormous savings In every in-terview and through our calculations there is nodoubt that a well-handled transfer of an activity cangenerate at least a 40 savings on the entire pro-cess (not just on labor costs) Table 3 presents thecost differences between operating a call center inKansas City Missouri in the United States and a callcenter in Bangalore India One rm claimed thattheir savings over a 2-year period were as high as80 for some processes The evidence we werepermitted to see but not reproduce suggests

strongly that quality is as good as or better thanthat achieved in comparable US facilities

These savings and quality claims cannot be inde-pendently veried however during the last 3 yearsthe overall growth rate in employment of BP opera-tions has been more than 50 per annum More-over every MNC we interviewed was dramaticallyincreasing the size of its Indian operations andmany said that the bottleneck was their ability togrow even more quickly The larger Indian independ-ents and Indian IT industry captives were growing ata similar pace Finally many other newcomers werebuilding facilities or exploring India with the aim ofestablishing operations This is prima facie evidencethat there are signicant economic benets Havingsaid that we are careful not to assume that this ratewill continue indenitely We identied several fac-tors that must be considered before concluding thata particular business process can be offshored in-cluding the complexity of the task the level ofinteractivity its knowledge component and level ofseparability the possibility of reengineering the ad-vantages of a single location and of scale and thetime-sensitive nature of the task However thereseems to be an enormous opportunity for furthergrowth when one considers that in 2002 more than85 million Americans were in the service sector (USBureau of Labor Statistics 2003) If only 2 of thoseservice jobs could be transferred this would be 17million jobs and the savings would be enormousOur opinion is that 2 is not an unreasonableestimate

In this study we identied seven types of BP op-erations in India There is no reason to believe thatonly one type can survive At the present the MNCcaptives have the greatest number of employeesundertake the highest-value-added work and areable to get the greatest usage of each of their seatsControlling the process internally mitigates the risksthat can come from outsourcing Finally it allowsthe MNC to capture all the prots of offshoringMost MNC outsourcers should also succeed in Indiaas they will be able to provide their developed na-tion customers with security and with their globalpresence a guarantee of redundancy Their greatestdanger is that BP offshoring to India becomes soroutinized that they will have to compete with thelower-cost Indian outsourcers directly In general the

34 Information Technologies and International Development

Lift and Shift

14 For a recent discussion of globalization in various manufacturing industries see Kenney and Florida (2004)

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

ing signicant value addition For example a pub-lishing rm that initially only prepared drawings forchemistry texts now offers a full range of back-ofce services including copyediting formattingand technical support The enhanced capabilityallows not only the addition of greater value butalso provides greater bargaining capacity

An example of such capability development is theMumbai rm Kale Consultants which specializes inproviding services to the airline industry OriginallyKale offered specialized airline software packageshowever in 2000 it extended its offering to includeBPs Coupling of specialized proprietary softwaretools with BP outsourcing operations meant it couldoffer a more comprehensive package For custom-ers this created an incentive to use Kale and in themeantime created a more permanent or sticky rela-tionship (personal interview 2003)

Developing domain expertise and becoming aspecialist are difcult and they have risks becausethe rm becomes dependent on a single industry oractivity And yet they also offer the potential to oc-cupy niches that may not be drawn into the extremelyferocious competition in highly commoditized sectors

A large number of Indian-owned and operated rmshave been established for the sole purpose of offer-ing BP outsourcing services to foreign rms A num-ber of these are venture capital-supported and wereformed during the Internet boom with the objectiveof providing back ofce services to US Internetrms Not surprising the collapse of the dot-comsforced these Indian service rms to rethink theirstrategies Because these rms were supplying back-ofce work such as answering e-mails and Web-related questions it was not difcult to switch theirservice offerings toward the much larger voice sec-tor Still other independents were funded by venturecapitalists as part of the enormous hype and excite-ment about BP outsourcing to India

These independents are often dependent on afew larger customers making them vulnerable tothe termination of a contract An important exampleof this is EXL Service which was the largest Indianindependent BP outsourcing rm until October2002 At that time 90 of its revenue (S Arora2002) and more than 800 of its seats were dedi-cated to the US insurance rm Conseco However

in December 2002 Conseco led for bankruptcy andthe employees dedicated to Conseco dwindled to175 by April 2003 (Verma 2003) Because of thisEXL Servicersquos growth stalled while other rmssurged by mid-2003 however EXL Service had re-covered and was once again growing

The strategic difculties are signicant Becauseof the ferocious competition and the felt necessityto expand they are under pressure to pursue manybusiness prospects However this mitigates againsttheir expressed desire to develop domain expertiseFor example the call-center-oriented rms must de-cide whether they want to specialize in in-bound orout-bound callingmdashtwo different skill sets Anotherdifculty is that the US market is the largest in theworld but sizing a facility for the US market meansthat during the day in India the facility is often idleThe independents have been able to secure somebusiness from Europe (especially England) that al-lows them to extend facility utilization however itis still difcult to use the entire facility for more than15 shifts One method for securing greater capacityutilization is to secure activities that do not requirereal-time processing12

The ultimate fate of the independents is difcultto predict and for the smaller independentsrsquo survivalwill be precarious Their greatest difculty is in mar-keting their services to foreign rms The larger in-dependents should be able to strengthen their USmarketing thus increasing their market share Themiddle-tier independents might be acquired eitherby Indian rms or multinationals wishing to enterquickly the BP outsourcing eld The strongest inde-pendents may be able to create rms that resemblethe multinational independents

The Indian IT industry has grown remarkably rapidlyover the last decade through the provision ofoutsourced programming and IT services to theglobal market (A Arora and Athreye 2002 DrsquoCosta2003) In IT outsourcing Indian rms such as HCLInfosys Satyam TCS and Wipro have become glob-ally competitive Because of their ability to uselower-cost Indian software talent they have madesignicant global market share gains Furthermoretheir interaction with the global economy has con-tributed to the development of executive and mana-gerial talent capable of securing overseas contracts

32 Information Technologies and International Development

Lift and Shift

12 This section is based on interviews with Indian executives

managing the interface with foreign customers andmigrating activities across national and rm bound-aries In the process these rms have cultivatedclose connections with foreign customers for IT ser-vices This provides an entree and condence on thepart of customers that facilitates persuading foreigncustomers to trust the outsourcing rms with otherservices

Given the growth in ITES the Indian IT rms be-lieve BP outsourcing is a sector in which they can ex-pand Their strategic question has been how toenter this new industry The major rms have an-swered this question differently Infosys and Satyamestablished subsidiaries one of which Progeon hasgrown rapidly and recently split a 5-year $160 mil-lion contract from British Telecom with HCL BPoutsourcing In contrast the Satyam subsidiary hasexperienced only limited growth TCS the largestIndian software rm entered the BP outsourcingsector through a joint venture Finally Wipro andHCL entered the industry through acquisitionsWipro acquired a venture-funded Indian independ-ent Spectramind which has grown quickly HCLacquired the Northern Ireland call center subsidiaryof British Telecom though most of HCLrsquos BP out-sourcing growth has been in India13

The Indian IT rms have signicant advantages interms of an ability to invest linkages to customersand various other strengths However the BP out-sourcing business is different from IT In terms ofmarketing the customerrsquos key decision maker is notthe chief information ofcer or chief technicalofcer Usually BP outsourcing is sold to the variousresponsible divisions or departments Furthermorethe ultimate decision rests with the chief financialofcer or chief executive ofcer This means thatthe Indian rm must operate through differentchannels

The BP workforce is also different Whereas inthe IT sector the workforce largely consists of engi-neers in BPs the workersrsquo degrees are in commerceand social science Because BP outsourcing work of-ten requires direct interaction with customers theworkforce-salient skills are interpersonal rather thantechnical Furthermore many BPs are undertaken inreal time therefore errors and mistakes have an im-mediate impact Service-level agreements are tightlywritten and monitored in real time therefore prob-lems are exposed nearly immediately In software

mistakes can be rectied later Also BPs that requirecustomer interaction can be extremely stressfulputting a premium on skillful workforcemanagement

The ability of Indian IT rms to manage nontech-nical personnel in extremely price-competitive envi-ronments will be tested However any test of theirmanagerial prowess may not come until later as therapid market growth will initially ensure an appear-ance of success for most entrants Difculties mayremain hidden until growth slows though by thattime they may have built such close relationshipswith their customers that exit by the customers mayno longer be possible because of its depth and stick-iness There is also the possibility that the technicalskills within the IT parent could be used to automateaspects of the BP outsourcing process creating an-other level of value addition that would improveprotability This would also enable the IT rm sub-sidiaries to create advantages beyond routine labor-cost arbitrage

There is a host of other established Indian rms en-tering the BP outsourcing industry Attracted by theldquoGold Rushrdquo aspects of the sector these traditionalrms with their roots in the large Indian businessgroups have invested signicant sums Already someof them appear to be experiencing difculties in se-curing customers (personal interviews 2003) How-ever in contrast with the smaller independentsthese captives have deep pockets and can competefor as long as their parents are willing to providesubsidies They will either nd a successful strategyor they will exit the business because of an unwill-ingness of the parent to sustain further losses

This genre of captives is interesting because oftenthey have no particular advantages In almost allcases there are few synergies between the parentrsquosexisting businesses and the services they aim to pro-vide They nearly always have experienced manage-ment though their experience may be in therelatively protected domestic market Frequentlythey have minimal experience in interacting withforeign clients especially in terms of providing ser-vices The lack of inherent advantages beyond deeppockets means that these captives will have to buildcapabilities in the same way as the Indian independ-ents Their only signicant advantage will be the rel-

Volume 1 Number 2 Winter 2003 33

Dossani Kenney

13 In June 2003 BT announced that it was establishing a captive Indian call center

ative deep pockets of their parent rms thoughoddly enough this may inhibit their ability to evolveto market demands In other words protection fromthe vagaries of the market could contribute to aninability to learn from the market

During the next decade the big debate about glob-alization will no longer concentrate on manufactur-ing which is an increasingly small portion of ourlabor force14 The globalization of services will takemanufacturingrsquos place The constantly falling price ofbandwidth the increasing digitization of informa-tion and the drive to reengineer and modularizeservices are making it possible to lift and shift themoverseas We argue that the favored destination isIndia and that at this time the rate of growth inthis transfer is signicant

The movement of business processes offshore isnot new Until the 1990s the undertaking of busi-ness processes in developing nations existed but wasminimal This paper shows how this may be in theprocess of changing dramatically We have beencareful not to exaggerate the current state of BPoffshoring Most rms have been cautious in theirdecisions to transfer these activities (especially mis-sion-critical functions) offshore especially to nationssuch as India or the Philippines that have differentsocial and political arrangements and sometimesproblematic physical infrastructures There are alsointernal risks that come from separating various ac-tivities in a process and possible resistance from de-veloped country workforces Finally a botchedtransfer process can be traumatic to the rmrsquos cus-tomers and employees

If there are dangers in offshoring BPs there isalso the potential for enormous savings In every in-terview and through our calculations there is nodoubt that a well-handled transfer of an activity cangenerate at least a 40 savings on the entire pro-cess (not just on labor costs) Table 3 presents thecost differences between operating a call center inKansas City Missouri in the United States and a callcenter in Bangalore India One rm claimed thattheir savings over a 2-year period were as high as80 for some processes The evidence we werepermitted to see but not reproduce suggests

strongly that quality is as good as or better thanthat achieved in comparable US facilities

These savings and quality claims cannot be inde-pendently veried however during the last 3 yearsthe overall growth rate in employment of BP opera-tions has been more than 50 per annum More-over every MNC we interviewed was dramaticallyincreasing the size of its Indian operations andmany said that the bottleneck was their ability togrow even more quickly The larger Indian independ-ents and Indian IT industry captives were growing ata similar pace Finally many other newcomers werebuilding facilities or exploring India with the aim ofestablishing operations This is prima facie evidencethat there are signicant economic benets Havingsaid that we are careful not to assume that this ratewill continue indenitely We identied several fac-tors that must be considered before concluding thata particular business process can be offshored in-cluding the complexity of the task the level ofinteractivity its knowledge component and level ofseparability the possibility of reengineering the ad-vantages of a single location and of scale and thetime-sensitive nature of the task However thereseems to be an enormous opportunity for furthergrowth when one considers that in 2002 more than85 million Americans were in the service sector (USBureau of Labor Statistics 2003) If only 2 of thoseservice jobs could be transferred this would be 17million jobs and the savings would be enormousOur opinion is that 2 is not an unreasonableestimate

In this study we identied seven types of BP op-erations in India There is no reason to believe thatonly one type can survive At the present the MNCcaptives have the greatest number of employeesundertake the highest-value-added work and areable to get the greatest usage of each of their seatsControlling the process internally mitigates the risksthat can come from outsourcing Finally it allowsthe MNC to capture all the prots of offshoringMost MNC outsourcers should also succeed in Indiaas they will be able to provide their developed na-tion customers with security and with their globalpresence a guarantee of redundancy Their greatestdanger is that BP offshoring to India becomes soroutinized that they will have to compete with thelower-cost Indian outsourcers directly In general the

34 Information Technologies and International Development

Lift and Shift

14 For a recent discussion of globalization in various manufacturing industries see Kenney and Florida (2004)

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

managing the interface with foreign customers andmigrating activities across national and rm bound-aries In the process these rms have cultivatedclose connections with foreign customers for IT ser-vices This provides an entree and condence on thepart of customers that facilitates persuading foreigncustomers to trust the outsourcing rms with otherservices

Given the growth in ITES the Indian IT rms be-lieve BP outsourcing is a sector in which they can ex-pand Their strategic question has been how toenter this new industry The major rms have an-swered this question differently Infosys and Satyamestablished subsidiaries one of which Progeon hasgrown rapidly and recently split a 5-year $160 mil-lion contract from British Telecom with HCL BPoutsourcing In contrast the Satyam subsidiary hasexperienced only limited growth TCS the largestIndian software rm entered the BP outsourcingsector through a joint venture Finally Wipro andHCL entered the industry through acquisitionsWipro acquired a venture-funded Indian independ-ent Spectramind which has grown quickly HCLacquired the Northern Ireland call center subsidiaryof British Telecom though most of HCLrsquos BP out-sourcing growth has been in India13

The Indian IT rms have signicant advantages interms of an ability to invest linkages to customersand various other strengths However the BP out-sourcing business is different from IT In terms ofmarketing the customerrsquos key decision maker is notthe chief information ofcer or chief technicalofcer Usually BP outsourcing is sold to the variousresponsible divisions or departments Furthermorethe ultimate decision rests with the chief financialofcer or chief executive ofcer This means thatthe Indian rm must operate through differentchannels

The BP workforce is also different Whereas inthe IT sector the workforce largely consists of engi-neers in BPs the workersrsquo degrees are in commerceand social science Because BP outsourcing work of-ten requires direct interaction with customers theworkforce-salient skills are interpersonal rather thantechnical Furthermore many BPs are undertaken inreal time therefore errors and mistakes have an im-mediate impact Service-level agreements are tightlywritten and monitored in real time therefore prob-lems are exposed nearly immediately In software

mistakes can be rectied later Also BPs that requirecustomer interaction can be extremely stressfulputting a premium on skillful workforcemanagement

The ability of Indian IT rms to manage nontech-nical personnel in extremely price-competitive envi-ronments will be tested However any test of theirmanagerial prowess may not come until later as therapid market growth will initially ensure an appear-ance of success for most entrants Difculties mayremain hidden until growth slows though by thattime they may have built such close relationshipswith their customers that exit by the customers mayno longer be possible because of its depth and stick-iness There is also the possibility that the technicalskills within the IT parent could be used to automateaspects of the BP outsourcing process creating an-other level of value addition that would improveprotability This would also enable the IT rm sub-sidiaries to create advantages beyond routine labor-cost arbitrage

There is a host of other established Indian rms en-tering the BP outsourcing industry Attracted by theldquoGold Rushrdquo aspects of the sector these traditionalrms with their roots in the large Indian businessgroups have invested signicant sums Already someof them appear to be experiencing difculties in se-curing customers (personal interviews 2003) How-ever in contrast with the smaller independentsthese captives have deep pockets and can competefor as long as their parents are willing to providesubsidies They will either nd a successful strategyor they will exit the business because of an unwill-ingness of the parent to sustain further losses

This genre of captives is interesting because oftenthey have no particular advantages In almost allcases there are few synergies between the parentrsquosexisting businesses and the services they aim to pro-vide They nearly always have experienced manage-ment though their experience may be in therelatively protected domestic market Frequentlythey have minimal experience in interacting withforeign clients especially in terms of providing ser-vices The lack of inherent advantages beyond deeppockets means that these captives will have to buildcapabilities in the same way as the Indian independ-ents Their only signicant advantage will be the rel-

Volume 1 Number 2 Winter 2003 33

Dossani Kenney

13 In June 2003 BT announced that it was establishing a captive Indian call center

ative deep pockets of their parent rms thoughoddly enough this may inhibit their ability to evolveto market demands In other words protection fromthe vagaries of the market could contribute to aninability to learn from the market

During the next decade the big debate about glob-alization will no longer concentrate on manufactur-ing which is an increasingly small portion of ourlabor force14 The globalization of services will takemanufacturingrsquos place The constantly falling price ofbandwidth the increasing digitization of informa-tion and the drive to reengineer and modularizeservices are making it possible to lift and shift themoverseas We argue that the favored destination isIndia and that at this time the rate of growth inthis transfer is signicant

The movement of business processes offshore isnot new Until the 1990s the undertaking of busi-ness processes in developing nations existed but wasminimal This paper shows how this may be in theprocess of changing dramatically We have beencareful not to exaggerate the current state of BPoffshoring Most rms have been cautious in theirdecisions to transfer these activities (especially mis-sion-critical functions) offshore especially to nationssuch as India or the Philippines that have differentsocial and political arrangements and sometimesproblematic physical infrastructures There are alsointernal risks that come from separating various ac-tivities in a process and possible resistance from de-veloped country workforces Finally a botchedtransfer process can be traumatic to the rmrsquos cus-tomers and employees

If there are dangers in offshoring BPs there isalso the potential for enormous savings In every in-terview and through our calculations there is nodoubt that a well-handled transfer of an activity cangenerate at least a 40 savings on the entire pro-cess (not just on labor costs) Table 3 presents thecost differences between operating a call center inKansas City Missouri in the United States and a callcenter in Bangalore India One rm claimed thattheir savings over a 2-year period were as high as80 for some processes The evidence we werepermitted to see but not reproduce suggests

strongly that quality is as good as or better thanthat achieved in comparable US facilities

These savings and quality claims cannot be inde-pendently veried however during the last 3 yearsthe overall growth rate in employment of BP opera-tions has been more than 50 per annum More-over every MNC we interviewed was dramaticallyincreasing the size of its Indian operations andmany said that the bottleneck was their ability togrow even more quickly The larger Indian independ-ents and Indian IT industry captives were growing ata similar pace Finally many other newcomers werebuilding facilities or exploring India with the aim ofestablishing operations This is prima facie evidencethat there are signicant economic benets Havingsaid that we are careful not to assume that this ratewill continue indenitely We identied several fac-tors that must be considered before concluding thata particular business process can be offshored in-cluding the complexity of the task the level ofinteractivity its knowledge component and level ofseparability the possibility of reengineering the ad-vantages of a single location and of scale and thetime-sensitive nature of the task However thereseems to be an enormous opportunity for furthergrowth when one considers that in 2002 more than85 million Americans were in the service sector (USBureau of Labor Statistics 2003) If only 2 of thoseservice jobs could be transferred this would be 17million jobs and the savings would be enormousOur opinion is that 2 is not an unreasonableestimate

In this study we identied seven types of BP op-erations in India There is no reason to believe thatonly one type can survive At the present the MNCcaptives have the greatest number of employeesundertake the highest-value-added work and areable to get the greatest usage of each of their seatsControlling the process internally mitigates the risksthat can come from outsourcing Finally it allowsthe MNC to capture all the prots of offshoringMost MNC outsourcers should also succeed in Indiaas they will be able to provide their developed na-tion customers with security and with their globalpresence a guarantee of redundancy Their greatestdanger is that BP offshoring to India becomes soroutinized that they will have to compete with thelower-cost Indian outsourcers directly In general the

34 Information Technologies and International Development

Lift and Shift

14 For a recent discussion of globalization in various manufacturing industries see Kenney and Florida (2004)

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

ative deep pockets of their parent rms thoughoddly enough this may inhibit their ability to evolveto market demands In other words protection fromthe vagaries of the market could contribute to aninability to learn from the market

During the next decade the big debate about glob-alization will no longer concentrate on manufactur-ing which is an increasingly small portion of ourlabor force14 The globalization of services will takemanufacturingrsquos place The constantly falling price ofbandwidth the increasing digitization of informa-tion and the drive to reengineer and modularizeservices are making it possible to lift and shift themoverseas We argue that the favored destination isIndia and that at this time the rate of growth inthis transfer is signicant

The movement of business processes offshore isnot new Until the 1990s the undertaking of busi-ness processes in developing nations existed but wasminimal This paper shows how this may be in theprocess of changing dramatically We have beencareful not to exaggerate the current state of BPoffshoring Most rms have been cautious in theirdecisions to transfer these activities (especially mis-sion-critical functions) offshore especially to nationssuch as India or the Philippines that have differentsocial and political arrangements and sometimesproblematic physical infrastructures There are alsointernal risks that come from separating various ac-tivities in a process and possible resistance from de-veloped country workforces Finally a botchedtransfer process can be traumatic to the rmrsquos cus-tomers and employees

If there are dangers in offshoring BPs there isalso the potential for enormous savings In every in-terview and through our calculations there is nodoubt that a well-handled transfer of an activity cangenerate at least a 40 savings on the entire pro-cess (not just on labor costs) Table 3 presents thecost differences between operating a call center inKansas City Missouri in the United States and a callcenter in Bangalore India One rm claimed thattheir savings over a 2-year period were as high as80 for some processes The evidence we werepermitted to see but not reproduce suggests

strongly that quality is as good as or better thanthat achieved in comparable US facilities

These savings and quality claims cannot be inde-pendently veried however during the last 3 yearsthe overall growth rate in employment of BP opera-tions has been more than 50 per annum More-over every MNC we interviewed was dramaticallyincreasing the size of its Indian operations andmany said that the bottleneck was their ability togrow even more quickly The larger Indian independ-ents and Indian IT industry captives were growing ata similar pace Finally many other newcomers werebuilding facilities or exploring India with the aim ofestablishing operations This is prima facie evidencethat there are signicant economic benets Havingsaid that we are careful not to assume that this ratewill continue indenitely We identied several fac-tors that must be considered before concluding thata particular business process can be offshored in-cluding the complexity of the task the level ofinteractivity its knowledge component and level ofseparability the possibility of reengineering the ad-vantages of a single location and of scale and thetime-sensitive nature of the task However thereseems to be an enormous opportunity for furthergrowth when one considers that in 2002 more than85 million Americans were in the service sector (USBureau of Labor Statistics 2003) If only 2 of thoseservice jobs could be transferred this would be 17million jobs and the savings would be enormousOur opinion is that 2 is not an unreasonableestimate

In this study we identied seven types of BP op-erations in India There is no reason to believe thatonly one type can survive At the present the MNCcaptives have the greatest number of employeesundertake the highest-value-added work and areable to get the greatest usage of each of their seatsControlling the process internally mitigates the risksthat can come from outsourcing Finally it allowsthe MNC to capture all the prots of offshoringMost MNC outsourcers should also succeed in Indiaas they will be able to provide their developed na-tion customers with security and with their globalpresence a guarantee of redundancy Their greatestdanger is that BP offshoring to India becomes soroutinized that they will have to compete with thelower-cost Indian outsourcers directly In general the

34 Information Technologies and International Development

Lift and Shift

14 For a recent discussion of globalization in various manufacturing industries see Kenney and Florida (2004)

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

specialists Indian and MNC that have created de-fendable niches should also be successful The inde-pendent Indian outsourcers will nd it more difcultThere are already rumors that the smaller out-sourcers that is those with insufcient capital haveclosed their operations The larger outsourcers thathave developed good customer relations and havesufcient cash ow to support marketing operationsabroad should be more successful though some ofthem are likely to be acquired The IT rm subsidiar-ies also may have a mixed prognosis The larger sub-sidiaries that have a good and growing customerbase are likely to do well through leveraging thedeep pockets contacts and expertise of their parentrms Finally we believe that the subsidiaries of theIndian business groups are likely to fail Thus themost likely future conguration of the Indian indus-try will be a complicated mix of Indian rms andMNCs Within this mix it is likely that the MNCs willconcentrate on the highest value-added operations

Regardless of the industrial structure it is nearlycertain that in terms of its contribution to employ-ment and value addition the BP industry will soonovertake software The industry also provides newcareer paths for Indian social science and commercecollege graduates who earlier faced high unemploy-ment rates In contrast with software which contin-ues to be heavily concentrated in Bangalore andMumbai there is evidence that the industry isquickly diversifying its locations to the large second-tier cities This should spread the employmentbenets to a larger number of young Indians For In-dia BP offshoring appears to have few detractionsbeyond the fact that wages may rise in the mediumterm

One aspect of BP offshoring that we have notexplored in detail in this paper but that is worthy of

further study is the rapidity with which it mightoccur Manufacturingrsquos movement offshore was agradual migration that has been under way since atleast the early 1960s Though punctuated by dra-matic factory closings there was ample opportunityfor the US economy to adjust15 This may not betrue in services where the objects are pixels andelectronic pulses that can be transmitted by photonsand radio waves (Cohen Zysman and Delong 2000Kenney 1997)

The impact on developed nations is more difcultto predict In the case of the United States even if2 of the service jobs or 17 million were lost theJuly 2003 unemployment rate of 62 would onlyrise to 73 Moreover these losses would not oc-cur instantly but would be spread over several yearsWhat this indicates is that the movement of servicejobs will likely provide some downward pressure onwages and the employment rate but will not be dra-matic In the longer run if it is possible to move sig-nicantly more business processes and other servicesoverseas it might be possible that India will do toservices what China has done to manufacturing Thiswould be a far more fundamental shift but it is tooearly to state denitively that this is the future

Aron R and J Singh 2002 The Rush to Send Back-Ofce Business Overseas Retrieved October 162003 from knowledgewhartonupenn edu100902_ss1html

Arora A and S Athreye 2002 ldquoThe Software In-dustry and Indiarsquos Economic DevelopmentrdquoInformation Economics amp Policy 14(2)253ndash 273

Arora S 2002 October 21 ldquoEXL Plans to De-RiskBusiness Modelldquo Computer Express Retrieved

Volume 1 Number 2 Winter 2003 35

Dossani Kenney

Table 3 A Cost Comparison between Call Centers Operated in Mumbai and Kansas City2002

Kansas City 006 014 1000 510 1530

Mumbai 012 021 150 183 366

15 We are entirely cognizant of the hardships that were experienced by displaced workers abandoned communitiesand small rms unable to afford overseas operations We are also aware that some workers communities and rmswere unable to adjust

October 16 2003 from wwwexpresscomputeronlinecomarchivesshtml

Arthur W B 1994 Increasing Returns and Path De-pendence in the Economy Ann Arbor Universityof Michigan Press

Baldwin C and L Clark 2000 Design Rules Vol-ume 1 The Power of Modularity CambridgeMA MIT Press

Bell D 1973 The Coming of Post-Industrial SocietyA Venture in Social Forecasting New York BasicBooks

Callaghan G P Thompson and C Warhurst 2001ldquoIgnorant Theory and Knowledgeable WorkersInterrogating the Connections Between Knowl-edge Skills and Servicesrdquo Journal of Manage-ment Studies 38(7)923ndash 942

Cohen S and J Zysman 1987 Manufacturing Mat-ters The Myth of the Post-Industrial EconomyNew York Basic Books

Cohen S S J Zysman and B J Delong 2000ldquoTools for Thought What Is New and ImportantAbout the lsquoE-Conomyrsquordquo Berkeley Roundtable onthe International Economy BRIE Working paper138

Cole R E 1994 ldquoReengineering the CorporationA Review Essayrdquo Quality Management Journal(July)77ndash85

Curry J 1997 ldquoThe Dialectic of Knowledge-in-Production Value Creation in Late Capitalism andthe Rise of Knowledge-Centered ProductionrdquoElectronic Journal of Sociology 2(March)

Davenport T H 1993 Process Innovation BostonHarvard Business School

David P 1986 ldquoClio and the Economics ofQWERTYrdquo American Economic Review Proceed-ings 75332ndash 337

DrsquoCosta A P 2003 ldquoUneven and Combined Devel-opment Understanding Indiarsquos Software Ex-portsrdquo World Development 31(1)211ndash 226

Deloitte Research 2003 On the Cusp of a Revolu-tion How Offshoring Will Transform the FinancialServices Industry Retrieved July 13 2003 fromwwwdccomInsightsresearchnancialoffshoringasp

Dossani R 2002 Telecommunications Reform in In-dia Westport CT Greenwood Press

Dossani R and M Kenney 2002 ldquoCreating an En-vironment for Venture Capital in Indiardquo WorldDevelopment 30(2)227ndash 253

Engardio P A Bernstein and M Kripalani 2003February 3 ldquoThe New Global Job Shiftrdquo BusinessWeek Available at wwwbusinessweekcom5A4e1YUQZcjedhMAmagazinecontent03_05b3818001htm

Feenstra R C 1998 ldquoIntegration of Trade and Dis-integration of Production in the Global Econ-omyrdquo Journal of Economic Perspectives12(4)31ndash 50

Geref G 1994 ldquoThe Organization of Buyer-DrivenGlobal Commodity Chains How US RetailersShape Overseas Production Networksrdquo inG Geref and M Korzeniewicz eds Commod-ity Chains and Global Capitalism (pp 95ndash 122)Westport CT Praeger

Hammer M 1990 ldquoReengineering Work Donrsquot Au-tomate Obliteraterdquo Harvard Business Review(JulyndashAugust)104ndash 112

Hammer M and J Champy 1993 Reengineeringthe Corporation A Manifesto for Business Revo-lution New York HarperCollins

Kenney M 1997 ldquoValue Creation in the Late20th Century The Rise of the KnowledgeWorkerrdquo in J Davis T Hirshl and M Stack edsCutting Edge Technology Information Capital-ism and Social Revolution (pp 87ndash 102) LondonVerso

Kenney M and R Florida eds 2004 LocatingGlobal Advantage Stanford CA Stanford Uni-versity Press

Kogut B and U Zander 1992 ldquoKnowledge of theFirm Combinative Capabilities and the Replica-tion of Technologyrdquo Organization Science3383ndash 397

NASSCOM-McKinsey 2003 NASSCOM-McKinseyReport 2002 New Delhi NASSCOM

OrsquoRiain S 2004 The Politics of High Tech GrowthDevelopmental Network States in the GlobalEconomy Cambridge UK Cambridge UniversityPress

36 Information Technologies and International Development

Lift and Shift

Orlikowski W J 1996 ldquoImprovising OrganizationalTransformation Over Time A Situated ChangePerspectiverdquo Information Systems Research7(1)63ndash 92

The Outsourcing Institute 2003 Retrieved July 152003 from wwwoutsourcingcom

Pearson R 1993 ldquoGender and New Technology inthe Caribbeanrdquo in J Momsen ed Women andChange in the Caribbean (pp 287ndash 295)Bloomington IN Indiana University Press

Porter M E 1990 The Competitive Advantage ofNations New York Free Press

Posthuma A 1987 ldquoThe Internationalization ofClerical Work A Study of Offshore Work Servicesin the Caribbeanrdquo Occasional Paper Science Pol-icy Research Unit University of Sussex Brighton

Reich R 1991 The Work of Nations New YorkKnopf

Smith A 2003 Wealth of Nations New York Ban-tam [Originally published 1776]

Sturgeon T J 2002 ldquoModular Production Net-works A New American Model of IndustrialOrganizationrdquo Industrial and Corporate Change11451ndash 496

Sykes Enterprises 2003 Sykes Enterprises Incorpo-rated Reports Fourth Quarter and Full Year 2002Financial Results Retrieved August 22 2003from wwwcorporate-irnetireyeir_sitezhtmltickerSYKEampscript5410amplayout5-6ampitem_id5

380769

US Bureau of Labor Statistics 2003 Retrieved onAugust 2 2003 fromftpftpblsgovpubsupplempsitceseeb1txt

Verma P 2003 April 17 ldquoEXL Service to DoubleStaff Countrdquo Financial Express Retrieved Octo-ber 16 2003 from wwwnancialexpresscomfe_full_storyphpcontent_id=32368

Warf B 1995 ldquoTelecommunications and theChanging Geographies of Knowledge Transmis-sion in the Late 20th Centuryrdquo Urban Studies32361ndash 378

Volume 1 Number 2 Winter 2003 37

Dossani Kenney