This Research Bulletin has been published in Environment and Planning A, 40 (2), (2008), 446-463 under the title 'Location Theory in Reverse? Location for Global Production in the IT Industry of Bangalore'.
Please refer to the published version when quoting the paper.
Contemporary urban theorists have highlighted the significance of growth of producer services in determining physical structure of the metropolitan areas of major cities in the world. As cities become integrated into the global economy, the services required by firms to navigate the complexities of a highly dispersed economy, agglomerate in specific locations in the city. Location studies of producer services have mostly concentrated on cities which are the control centers of the new economy. Little research has gone into studying location change in cities which are the receivers of global investment or more specifically the hosts of global outsourcing. These cities represent the other end of the production spectrum.
The following paper is aimed at developing a location model for a globally outsourced service industry in its host city. The specific industry under study is the Information Technology (IT) industry, owing to its similarity to producer services, which has agglomerated in the city of Bangalore in southern India. The paper is divided into eight sections. In the first section, the context of the globally outsourced Information Technology industry is established with respect to Bangalore. Next, the work done in the IT industry is compared with producer services to assess the relevance of existing knowledge of location to studying location change in Bangalore. Following this a brief introduction is made to the theorization of location of producer services in urban studies. In the fourth, fifth and sixth sections, the criterion for selection of firms for case study in Bangalore is presented, the overall metropolitan pattern of location discussed and finally a detailed account of location choices of case study firms is given. In the next two sections, findings from the case study firms in Bangalore are synthesised into an evolutionary model of location which is compared and contrasted with existing knowledge on intra metropolitan location of producer services. A number of initial conclusions, summarized in the final section, are then drawn on the basis of this analysis.
THE GLOBAL I.T. INDUSTRY AND BANGALORE
Developments in the Information Technology industry are often cited as the basis of contemporary globalisation – both in terms of technological advancements that enabled other economic processes as well as the global dispersal of production in the industry itself. Internationalisation of the semi conductor industry from the United States to East Asia during the 1960s has been the basis for a globally dispersed IT industry. The offshoring of production initiated by Fairchild Computers from the Silicon Valley to Hong Kong gradually spread to Taiwan, South Korea and Singapore (Henderson, 1989). What started as a dispersal of production of silicon chips and semiconductors gradually spread to Software and Services, together comprising Information Technology (IT). Though the ‘Asian Miracle’ cannot be attributed to one single factor, it is suggested that East Asia succeeded, in part, because it provided strategic, selective and consistent support to industries which were in the best position to facilitate their economies' transition from a labour-intensive to a knowledge-intensive industrial structure (Hanna et al, World Bank, 1996).
Success of the high tech production strategies of the East Asian economies acted as a demonstration effect for other developing countries as a strategy to ‘leapfrog’ into an information society without going through the process of routine industrial growth. The Human Development Report of 2001 mapped this new global network of production of Information Technology (IT) and Services to regions of technologically advanced production such as Campinas and São Paulo, Brazil; Bangalore, India; Kuala Lumpur, Malaysia; Gauteng, South Africa; and El Ghazala, Tunisia. These regions, majority of which are located in developing economies, have emerged as players in the global market.
The IT Industry that expanded rapidly in India after the 1991 economic reforms has attracted the attention of the global economy for its potential for expansion and rate of growth. “The industry generated 330 billion Rupees (USD 7.7 billion) in 1999, 15 times the level in 1990, and exports rose from USD 150 million in 1990 to nearly USD 4 billion in 1999. One study estimates that this could rise to USD 50 billion by 2008, leading information technology to account for 30% of India’s exports and 7.5% of its GDP. Employment in the software industry is projected to rise from 180,000 in 1998 to 2.2 million in 2008, to account for 8% of India’s formal employment” (HDR, 2001, p37). Bangalore City, located in the southern State of Karnataka in India has acquired international acclaim in the last 10 –15 years and is often know by its many sobriquets such as ‘Silicon City of India’, ‘IT Capital of India’, ‘Technology Hub’ etc. Its transformation from the ‘Garden City’ and the ‘Pensioners Paradise’ to the ‘Silicon City’ has been a steep curve of change effected in the last decade. The size of the export related IT industry in Bangalore has increased from 13 firms with an export output of 1.19 million USD in 1991-92 to 1322 firms and an export output of 4099.66 million USD in 2003-04. The composition of exports in 2003-04 consisted of Embedded and System Software, Enterprise Application Software and Telecom (76%), IT Enabled Services or Business Process Outsourcing (13%) and Others (1%) (all data from STPI, Bangalore, 2004).
With the first multinational firm (Texas Instruments) locating in Bangalore in 1984, it acted as a precursor to other such firms creating confidence in investors to locate in the city. Thereafter agglomeration economies were set in place for future investments. The nature of the industry changed gradually from research and development to execution of software designed elsewhere and most recently to back office services. As a result of the diversification, which is criticised by some as a ‘moving down the value chain’ (Saxenian, 2000; Parthasarathy, 2000), there is wide variety of firms located in the city. This fact is also illustrated by the composition of exports from the city. A significant development in the IT industry in Bangalore has been the growth of domestic firms such as Wipro, Infosys and Satyam Computers into large multinational firms as a result of the boom in offshore business coming to India. It is these firms which have moved up the value chain and have taken up end user application products and services.
LOCATION THEORY FOR THE IT INDUSTRY – CLASSIFICATION AS A PRODUCER SERVICE
Though the off shore IT industry is primarily associated with urban regions in Asia, it has received little academic attention in terms of location theory. This may have been due to the problem of its classification as a ‘service’ or an ‘industry’. Marshall and Woods (1995) have summarised the nature of service products as distinct from manufacturing by the following characteristics:
• services are ephemeral, lasting only for the period of any service transaction;
Further classifying services as ‘producer and consumer services’ Daniels (1985) has differentiated them on the basis of destination of their outputs (directly to consumers or to firms for further production) and durability (length of time over which service yields utility). Through examining the nature of economic activity in IT firms, some key characteristics can be identified which show their similarity with producer services -
• IT firms produce outputs which are utilised by other firms which may be from the service or manufacturing sector
Yet, there are aspects of dissimilarity with producer services - products of IT firms though satisfying the first two conditions put forward by Marshall and Woods (1995), can be owned and exchanged through the legal provisions of copyrighting and intellectual property rights. This to a certain extent creates a tangibility to their production. It is also not ephemeral or lasting only for the period of the service transaction. Even though producer services are exchanged among firms a certain level of face to face contact is required for the service transaction making proximity an important criteria for location. As opposed to that IT firms can undertake the execution of a project at a remote location after an initial agreement has been made by the firms. The person actually producing the software might have little or no contact with the end user of the product. The concept of ‘durability’ (Daniels, 1985) for producer services implies that the service interaction is one to one and the product cannot be used by multiple users. Whereas the software produced by IT firms is at another level of durability than for example business or management consultancy. The product in case of IT firms, though intangible, can be used by multiple users. In producer service firms, the service once produced is customised for the single client and would have only limited relevance to other firms.
While the similarity of production processes in IT firms permits their classification as ‘producer services’ the dissimilairites become significant when attributes determining location are being compared. Critical differences such as the limited need for face to face contacts in firms producing IT services for overseas clients necessitates a modified theory of location for this sector. However, in absence of such an existing body of knowledge, this paper builds on producer service location theory and is aimed at bringing to the fore variations from tendencies observed in producer services.
EXISTING KNOWLEDGE ON PRODUCER SERVICE LOCATION
Empirical academic interest in the physical distribution of services at an intrametropolitan and regional scale was in response to the growing significance of services as a part of the post war economy of the U.S. and Western Europe. Restructuring of the metropolitan economies had physical implications which centered around the congestion in central cities in the 1960s and a consequent suburbanisation from the 1970s. Hartshorm and Muller (1986) have suggested that urban America has gone through four stages – Bedroom Communities (pre 1960s), Regional Shopping Centres (1960 – 70), Growth of employment centres (1970 – 80) and the most recent devlopment of Suburban Downtowns (1980 – present). Of the four stages, the last three have been related with the suburbanisation of services, both household and producer. The suburbanisation of the latter form of services which were traditionally bound to the CBDs has been the most unexpected aspect of urban change in the last decade. Authors such as Stanback (1991), Cervero (1989) have referred to it as the ‘new suburbanisation’ of the ‘third wave’ which has lead to the formation of ‘edge cities’ (Garreau, 1991) and ‘suburban downtowns’ (Hartshorn and Muller, 1986 c.f. Champion, 2001, p150).
In addition to the suburbanisation of services, location change within the city and the city region has been theoriesed in the Multinucleation Model put forward by Daniels (1985). The model suggests an evolution of a multinuclear city pattern through a series of location and relocation by firms from the early 1960s to the 80s in urban areas of America. The process of relocation of producer service firms away from the city centre has resulted in a dispersal of back office functions to suburban location with major control functions remaining in the CBD. The ‘Seed Pod Model’ put forward by Schiller (2001) explains the dynamics of dispersal tendencies from the centre and suggests a cyclical process whereby firms initially locate in prime central locations, expand and then disperse to suburban locations as it becomes too expensive to expand in the CBD. The space vacated by dispersed firms is then taken up by other firms. However, even Schiller points out that certain critical functions such as finance, government, corporate HQs are retained in their original locations.
Empirical studies on the determinants of location for producer service firms can be grouped into three broad categories. The first, relate to role of communication and technology, second to organisational restructuring and its spatial repercussions and the last are related to other locational determinants (may be internal or external to firm).
In the first category are studies initiated in the 1970s aimed at evolving a comprehensive policy of relocation of offices in Western European cities and to explain suburbanisation in the U.S. cities. Prominent among these are works by Thorngren (1970), Goddard and Morris (1976), Daniels ed (1979), Pye (1977) and Gad (1979). In these studies of large cities in the 1970s, emphasis was on deciphering the contact patterns of firms and studying the loss in costs when they moved from central cities to the suburbs. The changes in communication as a result of the location change could only be compensated by increased travel or developing new contacts in the different location. Face to face contacts formed the key factor which bound offices to locations. But a parallel body of theory, initiated by the visionary writing of Meier (1962), discussed the possibilities of the substitutability of face to face communication with telecommunication. He suggested that given the developments in telecommunication i.e. video conferencing at that time, it would soon be possible to disperse activities away from cities and replace face to face contacts with audio visual conferencing. The 1980s brought both optimism and caution regarding the potential of telecommunications to replace face to face contacts. Daniels (1985) suggested that the adoption of new forms of information technology were expected to make the service sector vulnerable to ‘mechanisation of routine tasks’ (ibid, 1985, p. 273). A sequence of centralisation, fragmentation, dispersion, diffusion and ultimately the domestication of work places, as formulated by Niles et al (1976), would reconfigure firms such that ultimately home based workers would be linked to a hierarchy of offices with an advanced telecommunication network. Gottman (1983) cautioned against this ‘propoganda by the technologists’ and proposed that the adoption of technology is highly dependant on cultural factors and preference of individuals. Even though the new forms of communication offer unlimited possibilities for automated contact without physical movement, they can only reduce direct interaction and not eliminate the advantages offered by central cities.
Illeris (1996) discusses that the 1990s modified the approach to service location. While the view that distance friction does not vanish through improved telecommunications, was maintained, it is also being acknowledged that there is a time gap between the introduction and assimilation of new technology. As reduction in transport costs accompanied developments in communications, the decentralisation of activities has become more feasible. Even if services are dependent on high level of face to face contact, if they are high order producer services which are structured around high skills, physical transport forms a small part of the overall cost of the service and hence centralisation is not necessary. In comparison to this, individualised low and medium level services which are customised for clients and demanded at a higher frequency than more sophisticated producer services have greater need for physical proximity. The low per unit cost of these services requires the minimisation of transport and communication costs. The reduction in cost and time required for transfer of material and non material products (such as information which can be transported electronically) has brought into focus ‘other’ factors of location. It is no longer necessary for service providers to serve consumers in physical proximity. The concept of hierarchy of places being directly related to the level of services available is no longer a universal rule.
The second group of empirical studies concentrate on the influence of organisational restructuring in the era of globalisation on location of service firms. The shift from a Fordist to a Post Fordist way of organising production has had an impact on location needs of firms. Marshall and Woods (1996) state that service work has become more routinised and automated for lower order services while higher order producer services have expanded. As firms have slimmed their organisations to focus on core capabilities, the externalisation or outsourcing of sophisticated services has become important. The interdependence among the in-house production of business services and the external supply from service firms creates conditions for spatially agglomeration of such firms in large metropolitan areas. As Marshall and Wood highlight, offices of manufacturing firms have split their functions into Head Quarters and more routine functions. Organisational restructuring is further aided by the developments in information and communication technology. The major and most obvious impact of the developments in technology, which facilitate the instantaneous transfer of document information irrespective of distance, has been the increase in the physical range over which tasks of the service industry or any other economic sector can be performed. Illeris (1996) suggests that because firms are able to interact with other dispersed offices of all levels in the organisational hierarchy they are no longer bound to provide services/ products for the market in their physical proximity.
The third category includes ‘alternative’ approaches or ‘other’ determinants of location of firms that have been highlighted from empirical studies of specific places/ cities mainly after the 1990s (Coffey, 2000). These studies present factors other than communication which could be significant determinants of location especially with the flexibility afforded by new information technology. These ‘other’ determinants may be quality of life, taxations and regulation, life cycle of firms, initial contacts, image considerations of Metropolitan Philadelphia (Bondenman, 1998); site attributes and establishment attributes in Montreal (Coffey et al., 1996); dependance on sub contracting by federal agencies in Washington D.C. (Campbell and Harrington, 1997); historical forces and existing physical structure in Milan (Airoldi et al, 1997); creative milieus among advertising firms in New York (Leslie, 1997) and ethnic enclaves and connections among chinese producer service firms in Los Angeles (Zhou, 1998).
SELECTION OF FIRMS FOR STUDY AT INTRA-METROPOLITAN LEVEL
In order to get a wide mix of firms for the study of intra metropolitan location in Bangalore, the criterion adopted was based on a classification given by India’s National Association of Software & Service Companies (NASSCOM). The classification groups firms according to their share in the total software exports from the country.
Table 1: Segments for Case Study Selection (Adapted from NASSCOM, 2002)
* Calculated at exchange rate of 1USD= Rs. 47, # Included in two segements because of nature of production
As illustrated above, the firms for case studies have been selected as per the export segments in the existing pattern of the industry in the country. The 8 case studies vary according to the level of revenues, ownership, and types of products. They are all export oriented firms with the majority of the production being for an external market. In that sense they may all be considered ‘global firms’. All the firms have a considerable presence in Bangalore with their Head Quarters/ India Head Quarters located in the city. The oldest firm is Wipro Ltd. which began operation in the Software sector in 1983 and the newest being Acusis which began operation in late 2001.
For the purpose of clarity ‘firms’ in this paper will be distinguished by the nature of their production – Software firms include all such firms that undertake production of programming applications for end users or do production of applications/ parts of applications on contract by other firms. IT Enabled Service(ITES) firms are those firms that provide services enabled by communication technology for labour intensive back office operations. These operations may involve direct contact as in the case of Call Centers, Customer Service or may relate to back office functions for bookings, medical transcription etc.
LOCATION PATTERN OF THE IT INDUSTRY AT METROPOLITAN LEVEL
History of the IT industry in Bangalore in its present export oriented form can be traced to the mid 1980s. The firms as well as the business has since expanded and transformed with 4 major identifiable shifts brought about change in communication infrastructure, public policies and establishment of the credibility of India in the world market. The four phases of growth have been: Phase of initial MNC location (mid 1980 - 91), Phase of ‘body shopping’ firms (1991-96), Phase of Offshore Development Centres (1996-99), Phase of Product Development and ITES (1999-present).
Each of these phases has been characterised by location diversification in the city. In the first phase, the concentration of offices was in the CBD. This was the location where most office space was available and the rents were affordable to firms. In the second phase, when space was constrained in the CBD and the wireless technology used in the Satellite station required clear sight access for contact, dispersal of firms took place to the South and South East of the city. Secondary Business Districts (SBDs) also developed around the main CBD. In the third phase, after a crash in real estate prices and large over supply of space, high quality office space developed by the government in the form of an advanced Information Technology Park began to attract firms. Smaller firms which could not afford the high rents in the CBD, SBD and peripheral Technology Parks started to locate in residential properties which were converted to office space. Thereafter, the slowdown in the global business has made firms as well as developers more cautious. Large firms are still consolidating offices by opting for campuses in suburban locations. Developers are entering into more innovative lease agreements which ensure that space is leased and the construction is built to suit the needs of the lessee.
When the spatial typology of agglomeration of firms is analysed, four major areas of concentration can be identified each of which represents a unique urban pattern. These areas are:
• CBD and the SBDs
The CBD and SBDs are characterised by multipurpose office complexes which might be occupied by single or multiple tenants which are not necessarily IT firms. While the CBD is also a recreation and shopping destination, SBDs are primarily locations for offices and ancillary support services. Municipal areas in the south and east quadrant are planned residential neighbourhoods developed during the 1960s - 80s by the Bangalore Development Authority. Plots planned for single unit row housing have been modified by IT firms and are used as office space. The strip development along major highways leading out of the city mainly in the South and South East direction is a case of suburban agricultural land bought over by developers and converted to offices in the form of customised campuses for large IT firms. Special zones in the periphery are exclusive developments for the promotion of the IT industry and are linked with specialised incentives and infrastructure. Two existing Technology Parks have already been operational for the last 5-7 years and the State is in the process of extending the localised area of the Technology Parks in the form of an IT corridor. The following Figure 1 illustrates the location of areas of concentration of firms in the Bangalore Metropolitan Area.
Figure 1: Areas of Concentration of the IT Industry in Bangalore
Concentration of firms in the metropolitan area of the city, as well as the development of infrastructure for the IT industry has lead to two distict urban phenomena – these are Multinucleation and Peripheralisation of Offices and Retail Space. As firms seek out areas for expansion and dispersed location, the city has changed from a single core CBD to multiple commercial and retail cores. In addition to this, new and high quality development of infrastructure and commercial space (linked to investment incentive) by the State has resulted in increased property prices in the periphery. Land values and rents in these Technology Parks are at competitive levels of the CBD, leading to an increased peripheralisation of offices and retail activities. The following sections of the paper illustrate these processes of urban restructuring with respect to spatial decisions taken in the selected firms. The sequence of spatial decisions taken by the firms studied is used to formulate a theoretical model of location of IT firms at an intrametropolitan level.
EVIDENCE FROM SELECTED FIRMS – SEQUENCE OF INTRA METROPOLITAN LOCATION
From the eight selected firms, the sequence of location of only five are discussed here in detail.. Of these three are Software firms while two are ITES firms. Though the theoretical model proposed at a later stage incorporates spatial decisions in all the case study firms, they are excluded here owing to their shorter time span in the city and hence not displaying a diverse range of location change.
Spatial Decisions and Location Change in Software Firms
The first Software firm selected for discussion is Wipro Technologies Ltd. The firm belongs to large business group which diversified into Information Technology in 1983 with the setting up of a subsidiary. The city of Bangalore had been the Head Quarters of the business group for thirty years before the IT firm was set up. Hence, Bangalore was also its natural choice as Head Quarter for the new firm. Initially, the administrative office was located in the CBD and the Software Development Centers in three other Development Centers around the CBD. The period from 1985 - 95 was a testing phase for the firm and no major investment in property was made unless the certainty of growth in this sector was established. Till 1995, the firm had grown from 30 employees to almost 2000 employees. These were divided by 300-400 persons in the 4 offices in the CBD. The period from 1995-96 brought about a shift from purely Onsite to Offshore software production by Wipro. Once the credibility of the firm had been established clients were willing to totally outsource the projects and increase the Off shore element in the projects.
Since expansion in the CBD was restrained by the phenomenally high rental and lack of large floor space, the firm chose new Development Centers in the residential areas while maintaining the corporate office in the CBD. The first significant shift outside of the CBD was to two offices in a residential neighbourhood in the Municipal Area. As and when large projects were acquired by the firm, a new space was leased in the proximity of the earlier Development Centers and an ODC (Offshore Development Center) established. The ODC is a model of functioning in which the Offshore Office becomes an extension of the office of the client and all systems are set up as replicas of the client. The office thus is required to be segregated from the other project offices either physically or through electronic security systems. Therefore, Wipro developed a network of development centers in the city and chose even later not to consolidate them in a single physical location. Subsequently, as the IT industry boomed in 1996 and the firm had established its brand name, Wipro started a phase of investing in property in the city. They undertook two major projects - one, to own and develop the largest Development Center in the periphery in Special Economic Zones, and two, land was purchased along the proposed peripheral Ring Road for a consolidation of the Corporate Head Quarters of all the businesses in a single location. In late 2001, both projects were completed and the Corporate Head Quarters of all Wipro companies were moved to the peripheral location about 12 kms. from the Municipal Area boundary.
The cycle of location of the firm can be divided into the following three phases:
• Phase 1: Initial location of HQ and Development Centers in the CBD and Secondary CBDs in Commercial Complexes
The second Software firm is the wholly owned subsidiary of IBM International. IBM which has been in the city for the last 10 years has a number of large offices all over the country but still decided to locate its Head Quarters for the India operations in Bangalore. In 1993, when it started operation in Bangalore, the firm came in as a joint venture between the Tata Group (Large Indian business conglomerate) and IBM International. The firm then occupied an office space owned by the partner firm in a prime commercial complex near the Airport. This office remains until today the Registered Office of IBM India. In 1997, the stake of the Tata Group in IBM was reduced to 20% and IBM Global Services was extended to India. This was the time that software development was taken up in India and the firm leased two office spaces for Development Centers, one each in the CBD and the secondary CBD. In 1998, the ownership was transferred completely to IBM International and the firm became a fully owned subsidiary. At the same time, in order to expand the IBM research and development located in Bangalore, the firm made a decision to consolidate its space in the city. A long term lease agreement was made with a builder and landowner for a space about 10 kms. from the CBD along the corridors leading to the Special Economic Zones. The agreement involved the renovation of an existing building to suit the needs of the firm and a special agreement for future expansion in the land abutting the building. Though the firm had two options for sites for a consolidated campus, the second one being along an alternative corridor at a distance of about 15 kms. from the CBD, the current site was selected due to the easier terms of lease agreement which involves minimum permanent commitment and the choice of employees to stay closer to the city. After the crash of the IT global business, the expansion plans for the construction of a large campus on the site were put on the back burner, but the development centers were mostly moved out to this single consolidated space. Though the office near the Airport and the other two city offices are maintained as a back up space, the Development Centers and the Head Quarters are concentrated in the suburban location.
The shifting of location of the firm can be distinguished into three phases:
• Phase 1: Initial location near the Airport in leased office in a Commercial Complex
The last Software firm being discussed is I flex. I flex as an independent firm which broke away from the original firm called Citicorp (Subsidiary of MNC Citi- Group) was set up in Bangalore in 1993. From working exclusively on banking related software for the Citi-Corp, the firm was set up as a joint venture between the parent firm and capital raised by the former employees of Citi-Corp. The ownership of the firm still lies 40% with the Citi group but 1998 marked the final emergence of I flex Ltd. as an independent firm selling their banking related software. The original location was a leased office space in a prime commercial complex in the CBD. This office was already available to the firm as a part of Citi-Corp. However, with the expansion needs and the rapid growth of the firm by 1998, additional space for the second Development Center was leased in a Commercial Complex in the Secondary CBD. In 1998, the Head Quarters of the firm was also moved to Mumbai to achieve proximity to the Financial Institutions. In the next planned move of the firm, a site has been purchased in a suburban locality which lies in the East, near the Special Economic Zone. The construction of a 12,500 sq.m. building and campus has already been initiated and the scattered Development Centers as well as the Head Quarters is expected to be moved into the new campus by the end of 2003.
As with the previous two cases, the locational cycle of I flex leads to ultimate suburbanisation and a consolidation of activities in a single campus. Though Bangalore is not the corporate head quarter of the firm once the campus is constructed then Administrative Head Quarter will move to Bangalore. The phases can be distinguished as:
• Phase 1: Initial location in a commercial complex in the CBD
Spatial Decisions and Location Change in ITES Firms
Having discussed location decisions of 3 software producing firms, I discuss here the same for two ITES firms. There is a clear distinction of locational requirements for production of Software and Internet enabled Services. ITES firms mainly consider two factors that dictate a trade-off for location:
• Rent plus Transport cost/ employee
Among the Case Studies, First Ring and Acusis are both Service Firms which provide IT Enabled Services very much dependent on a high speed internet connectivity. The basic infrastructure for these kinds of firms is now available throughout the city with newly laid out Fibre Optic Network and yet they opt for totally opposite locations for their facilities. First Ring is located in the Technology Park almost 15 kms. from the Airport and Acusis is located near the CBD. Differences of locational choice exist between the firms in similar revenue segments and product profiles.
Comparing the locational choices of Acusis and First Ring, the influence of business models on locational choice is illustrated. Acusis which is a Medical Transcription firm relies on high speed communication to get voice data from its clients in the US, which are large hospitals, that has to be transcribed and sent back before the doctors come back to work the next day. The speed of communication and the transcribers is the key to sustaining the business. Similarly, First Ring is also a service firm which offers back office operations such as airline bookings, customer support, call centers etc. which have short turn around times. Both the firms have parent firms which are located in the US through which work is channelled to the main service centers located in India. However, First Ring, prioritising the quality of infrastructure, chose to locate itself in the Technology Park (20 kms. from the city center) and transports its employees working in 12 hour shifts from all over the city to this peripheral location. According to the Personnel Manager of First Ring, the cost of transporting the 900 odd employees is made up by the fact that the communication facilities available in Park are the best in Bangalore and enable high speed business.
On the other hand, Acusis has chosen to locate itself on Lal Bagh Road which is almost in the city center. The firm is based on a model of ‘Home Based Transcription’ (HBT) which enables the employees to get work at home through a high speed communication network and be able to transmit the output to the head office via the same network. The firm aims to employ about the same number of employees as First Ring by end 2003, but they would not be required to travel daily to the office thereby saving huge costs which have been channelled into having a smaller but centrally located office. The Finance Manager of the firm stated that the central location was selected to ensure that in case the employees involved in the HBT needed to contact the office, they could do so without having to travel large distances thereby saving time and money.
The contradiction in the above locational choices lies in the fact that one firm prioritising high quality communication networks opts for a high rent and high transport cost location whereas a similar firm opts to locate in a smaller and cheaper office in the city and employs technological solutions to minimise transport costs.
EVOLUTIONARY MODEL OF INTRA - METROPOLITAN LOCATION
Though there are differences among the firms and all of them are not at the same level of business development, a three phase process of location and relocation can be identified. It is represented in the following Figure 2.
Figure 2: Theoretical Model of Location Change in IT Firms
In the first phase, firms select an initial location in the city. The choice of initial location can be dependent on a number of factors. Among the cases studied initial location is dependent on the ownership of the firm and its existing size. Large MNC firms who locate initially to explore the market have an image as well as higher affordability as compared to domestic owned local Start Up firms. Image considerations in the market are not an important consideration for Start Up firms having just entered the market. The second factor which had an influence in the case of IBM as that of local partnerships and initial contacts. The office location in this case was dictated by the availability of space with the local partners. Not wanting to make immediate commitments in the city, the existing office space available with their partners was selected for initial location. I flex also occupied space available with its parent firm. The size of initial capital investment backing available to the firm is an obvious determinant of the nature and quantum of space selected for initial location. Established firms such as IBM or those backed by large industrial groups such as Wipro have more flexibility to select prime locations in the CBD than Start-ups such as Infoquark. For ITES firms, initial location is also a decision based on a tradeoff between quality of infrastructure and the cost of transport for the large number of employees for round the clock shifts. The small profit margins in this low value added sector are very sensitive to marginal cost increases of factors such as the speed of connectivity and the cost of transporting employees to destinations where the infrastructural quality is high.
In the second phase, firms expand their Development Centers and start to assume a multiple office patterns in the city. While the original office is retained and most control functions are not moved, the production centers or Development Centers are either expanded or multiplied. In some cases the control functions or the Head Office is also moved to a more prestigious location. This is the phase where a separation of functions within the firm starts to take place. Shift to the second phase takes place when the firms shifts from onsite production at the clients office to an offshore production. Offshore production means that project teams would be constantly stationed in the city and more space is required hence the expansion of Development Centers. Once the firm has established its reputation in market and clients can outsource projects completely in offshore destination then the employee strength of firms is increased. The increase in employee strength creates normal expansion needs adding to the space requirements. Increase in employee strength in the case of I flex and Wipro was also related to organisational reorganisation. As Wipro expanded its internal organisational structure was split into two separate divisions which were handling different markets. This created a need for multiple offices and hence the need for expansion.
Since firms create a specialised niche in the market, they sometimes handle projects of competing clients. For example, I flex which specialises in banking software or Wipro which is specialist in Enterprise Resource Planning (ERP) and E -Commerce, often have projects from firms which are competitors in their own consumer markets. To maintain the levels of project security required to attain client confidence a physical separation of work is required. Separation of projects from competing clients thus becomes and additional factor for maintaining physically dispersed Development Centers. This fosters the creation of a network of offices of single firms within the city. The dispersal is maintained either in physical terms or electronically if the Development Centers are consolidated at a later date.
In the third phase, firms consolidate control functions and some key Development Centers in user specific campuses while a dispersed pattern of Development Centers is maintained. Because only suburbs and peripheral locations offer the possibility of large campuses these are located outside the city with a high quality of firm owned infrastructure. A campus which was initiated by Infosys (second largest IT firm in India) has become a matter of prestige. Facilities such as large conference halls, video conferencing, employee entertainment facilities, firm owned satellite linkups, complete captive power backups, incubators etc. have become symbolic of a firm’s success. Thus upgraded infrastructure and image becomes an important factor in this third phase of firm relocation. The motivation behind the suburban and peripheral location for campuses is that large consolidated space is available at an affordable price. The suburbs offer a good environment and an option to escape from the over crowded central cities.
Firms such as Wipro and IBM which have reached this stage in their life cycle opt for user specified campuses which they do not necessarily own. Like IBM innovative lease agreements are possible with builders. The selection of location for the campus is based on a relative proximity to other Development Centres. To maintain the network of offices in the city, firms set up Local Area Networks and centralised servers where project information is saved and accessed. Despite that proximity adds to the convenience of movement between offices for face to face interactions. As reported by the Manager, Administration Facilities of Wipro Technologies, the functions within IT firms are divided into three areas i.e. technology, communication and enterprise. Technology areas are the Development Centres where actual production takes place. Communications is the functions which are required to manage interaction between dispersed offices and Enterprise is related to management and administrative functions. The consolidation and dispersal of these three categories of functions differs from firm to firm but in a relative consolidation takes place in Enterprise and key Communication and Technology functions in a campus.
The maintenance of the dispersed pattern in the city even though some functions are centralised is owing to the earlier discussed factor of maintaining project security as well as a second factor of obtaining maximum taxation benefits. Under the Software Technology Parks (STP) scheme of the Central Government individual premises are declared as 100% Export Oriented Units making them eligible for Sales Tax concessions. If a multiple office network is maintained projects can be moved around within the intra city network depending on the level of export segment in the output. Not all premises of firms are registered as STPs and hence are not obligated to achieve the stated goals for export. The logic of multiple offices in the city was also explained by a need to prevent disruption due to disasters. In view of the September 11 attacks in the U.S., firms have become conscious of the need to have multiple offices which are able to take over operations in case of destruction of any single office. Backups of key information and multiple links in networks are facilitated by the multiplicity of offices. This was a very significant factor in the case of IBM which has adopted this policy in its global offices. Wipro on the other hand adopted measures for ‘disaster recovery’ being under pressure from its clients.
The three phase process of location change observed among case study firms can be described as a dual process where firms have varying tendencies to both disperse from city cores and to centralise in other locations. The formation of secondary business districts within the municipal area as well as a concentration along traffic corridors and nodes in the periphery, though planned, shows the need to disperse in the former and centralise in the latter case. In a future scenario it is expected that as firms attain a certain size they would increasingly disperse from central city locations to campuses, technology parks and other peripheral real estate development.
However, not all firms among the case studies have gone through the three phases described above. Wipro and IBM which have both been in Bangalore for more than 10 years have shown all three stages. I flex is in the second phase and planning to relocate to a campus soon. Infoquark which is the newest Start Up is still only located in its first office. The factor of where the firm is placed in its life cycle is an important determinant of its pattern of location in the city. The ultimate peripheralisation and suburbanisation observed in firm location at city level is produced by firms which have had a certain level of growth over time in order for them to plan campuses and user specific, single owner occupied buildings.
Summarising the findings from the study it can be concluded that firms in the IT industry in Bangalore, are more influenced by factors which are internal to the firm rather than external influences, in determining location at an intra metropolitan level. However, a clearer distinction of factors with respect to existing theory can be made. The following Table divides the factors at each Phase of location in four categories - Communication factors, Organisational factors, Policy factors and Value preferences.
Communication factors relate to the considerations regarding physical transport and the substitutability of contacts with advanced technology. Organisational factors relate to changes and determinants which are related to the way a firm operates. Policy factors are those incentives and spatial policies which the firms avail of and are subject to in the urban context. Value preferences are socio cultural aspects of location such as the aspiration to a certain image associated with office space and preferences for a better urban environment.
Table 2 : Theoretical classification of determinants at each Phase of location
EXISITING KNOWLEDGE ON LOCATION OF PRODUCER SERVICES: PATTERNS OR DIFFERENCE?
Discussing implications of the evolutionary model suggested above and the determinants of location identified, with respect to existing knowledge on location of producer services, parallels can be drawn to the Multinucleation Model put forward by Daniels (1985) and the Seed pod model put forward by Schiller (2001).
Even though a similar pattern of multinucleation, phased shift of location and clustering in peripheral locations can be found among IT firms in Bangalore, a critical difference can be identified. Though firms choose initial locations in the city centre or other locations within the major urban fabric and they follow a path of expansion to finally consolidate in large campuses in the periphery, it is not the management and control functions which are retained in central locations. Unlike other producer services firms, IT firms do not have as many forward linkages to the local urban context and hence do not need to maintain proximity. It is the management and control functions, usually located in the Head Quarter which are consolidated in suburban and peripheral locations. Development Centres which require extensive and affordable space as well as proximity to residential areas in order to enable short commuting distance for the majority of the technical work force are retained in intermediate locations in the municipal area. Higher management staff and senior technical employees have higher mobility and hence can travel larger distances to suburban campuses where strategic control functions and selected Development Centres are consolidated.
The two main aspects of variation from existing theory on urban patterns of office location are: -
• The conversion of residential buildings and land to commercial use for these Development Centres in high income planned residential areas
In sum, the urban structure created is similar to those discussed in the Multinucleation Model and Seed Pod Models but the distribution of functions among the core city and suburban locations is different.
In contrast with existing theory, strategic functions of IT firms do not require to retain central city locations owing to their export oriented production. Firms which aim to target local markets still retain CBD or intermediate location offices. Other firms after attaining a certain size and having established themselves, seek to consolidate their strategic functions in campuses and single owner occupied buildings. However, even among IT firms, differences exist for IT service firms such as call centers and medical transcription. These firms having predominantly back office operation for a parent firm located in another country have even lesser need to locate in the CBD or even intermediate locations. Their decision for location is based on a trade off between rents and the cost of transporting their larger labour force for round the clock operation. Firms which prioritise larger offices with better communication infrastructure locate themselves in peripheral areas. The limited administrative and strategic functions are still bundled with the service production functions. On the other hand, a completely divergent location choice is made by firms which operate on home based workers and would like to minimise transport cost by locating a smaller but centralised office in the Secondary CBD.
In terms of the spatial form produced and the rent curves, high rise office development in the peripheral IT park has been a State led initiative and more related to establishing a certain image for the city. Private development in the suburban corridors and specialised peripheral centres is still low rise and low density. Land values in the peripheral office developments, though seem to converge with CBD prices. The spatial landscape of Bangalore shows a similar contrast of high rise buildings as compared with the low rise urban fabric. Peripheral high rise development and campuses show another level of contrasts from the surrounding agricultural land and informal urban development as that discussed in the account of American suburban growth but at a much different scale. Unlike European cities, Bangalore has no strict controls on redevelopment of the colonial CBD and conversion of old structures to high rise office blocks to gain maximum advantage of high land values is common. High rise and high land value in the periphery however, has been artificially created by the higher quality of space and infrastructure investment by public agencies. Planned infrastructure projects such as the Infotech Corridor are likely to increase the pressure on land in the periphery further and create a situation of suburban enclaves of high tech office spaces - a dispersed ‘global business district’.
In terms of determinants of location at each phase, existing theory pays great significance to the role of contact patterns of producer service firms (Daniels ed., 1979; Daniels, 1985; Goddard and Morris, 1976; Gad, 1979), their substitutability by developments in telecommunication (Daniels, 1985; Illeris, 1996; Beyers, 2000; Castells, 1989) and to a certain extent on organisational change (Dicken, 1998; Gottmann, 1983; Marshall and Woods, 1995; Illeris, 1996).
Goddard (1975) put forward the idea of dividing contacts into functional interdependencies (contacts between offices), spatial structure (contacts between employees within a firm) and physical movements of individuals and materials. The findings from the cases suggest that in export oriented IT firms contact within the firms (spatial structural contacts) and to certain extent physical movement of individuals (for ITES firms) form important factors in location decisions. In Phase II, Development Centers are dispersed to sever contacts among employees working in conflicting projects to maintain intellectual security whereas in Phase III a proximity of the strategic control functions and Development Centres is designed to maintain ease of physical movement between the offices. A balance between dispersal and proximity is preferred such that a physical separation is maintained but swift transport for necessary face to face contact for internal meetings is possible. However, functional interdependencies do not play an important role as other partner firms and clients are located all over the world. The work which is outsourced in Bangalore is routine operations of a larger multinational firm or specialised service outsourced by manufacturing firms, the need for direct regular contact is minimum. Of the office contacts identified by Thorngren (1973) as programmed (routine), planning (implementation of action) and orientation (strategic decisions to maintain competitiveness), programmed contacts with the parent firm are maintained through advanced telecommunications while less frequent planning and orientation contacts are undertaken by means of video conferencing or actual international travel. While analysing the relocation potential of firms in Stockholm, Thorngren suggested that functions of programmed and partly planning contacts could be decentralised. However, the developments of information and communication technology have enabled the decentralisation of programmed functions not only to suburban locations but even outsourced to global locations. Therefore for these firms which are mainly involved in programmed contacts with the parent firm and clients, intrametropolitan contacts are limited to maintain backward linkages and internal organisational contacts among employees of the same firm.
Reflecting on the role of technology in replacing face to face contacts, empirical studies suggest that with the increasing complexity in production networks which have tendency to be globally dispersed the significance of face to face contacts for strategic decisions is only increasing (Storper and Venables, 2002). Beyers (2000) finds that though means of communication and information exchange have multiplied, these only supplement the core role of direct contact. The findings from the case studies point to a confirmation of this aspect of intra and inter firm communication. Even though firms in Bangalore can disperse Development Centres in various locations in the city as well as globally, strategic management and decision making functions are consolidated and centralised in selected locations. Intra firm contacts with Development Centres are maintained through internet communications and information stored and distributed through centralised servers, yet the Head Quarters of large firms such as Wipro has been centralised further. Since the case study firms have limited forward linkages in the local urban area, contacts external to the firm are maintained through the use of communication technology such as e-mails, telephone, video conferencing and infrequent international travel by project teams and decision makers.
With respect to determinants of location (Refer Table 2) it can be seen that organisational factors form an important factor in location change of the case study firms. Theorists such as Dicken (1998) and Marshall and Woods (1995) have suggested that in the new ‘service economy’, organisations are more flexibly organised as are production processes. Transnational and Multinational firms that form the basic building block of a service economy and of post fordist production operate according to a large number of models (Dicken, 1998). Large IT firms like Wipro, are organised in a way that ‘each unit performs a separate part of the production sequence and units are linked across nations in a chain like sequence’. This is the model which forms the basis of operation at an intra city level as well. The most important factor of location change is the shift from ‘onsite to offsite production’ whereby a firm no longer sends project teams to the clients office but dispersed production centres are established which are responsible for disaggregating the production sequence. This creates the need in Phase II for spatially dispersed Development Centres which are linked to a centralised communication network. Therefore, it is organisational change which is responsible for expansion of the firm in a city and shift toward a ‘multiple office pattern’. Organisational changes such as mergers, acquisitions, break away firms etc. have also been factors behind change of location in the case study firms.
In Phase I and III, it is direct organisational demands that determine location. While in Phase I the affordability for prime space and access to space already available to partners are determined by the nature and ownership of the firm, in Phase III, firms size also determines the level of infrastructure required and the complexity of strategic decisions to be made. The choice for a campus which consolidates all management and strategic functions is made when a firm attains a certain size and requires dependable, secure and high level of infrastructure to manage its complex global business network. Other factors such as ‘image’ and other value preferences also play a part in Phase III.
Recent developments in theory find that as proximity constraints are relaxed dues to a large number of factors, ‘other’ determinants begin to play an important role in office location. From the findings of this research, there are two kinds of additional factors which play important roles in determining location - policy factors and value preferences. The multiple office pattern opted for by firms of relatively larger size which reach Phase II in the firm life cycle is related to the incentives offered for sales tax exemptions associated with specific registered production units. The incentives are also subject to export targets which the firms have to state and achieve annually. Hence firms maintain multiple offices such that they can allocate projects to units based on their export/ domestic market classification. In addition to this location policies and incentives of the State Government have had an impact on the selection of sites for campuses in the periphery as well as the supply of space by developers. Through planned developments such as the Information Technology Park where location is linked to infrastructure and fiscal incentives, agglomeration at a city level has been influenced.
But apart from that there are some value preferences such as ‘image’ and ‘better urban environment’ which firms make at advanced stages of their life cycle (mainly Phase III). While ‘image’ of the firm is a consideration for subsidiaries of MNCs, even at initial location, it is more related to affordability of prime space while smaller startups which are not backed by large investments or existing businesses are more involved in establishing their reputation in the market. Image becomes an important factor at Phase III when a campus or a single owner occupied office building is opted for. Image building through the campus is an important consideration for firms that have achieved a competitive position in the global market. Other value preferences such as a ‘better urban environment’ in peripheral locations also become important at this phase as the firm can rise above constraints of affordability. A suburban location where the firm can establish an office with secure and dependable infrastructure, recreational facilities, synergistic work environments and an image of success is an option exercised to escape the overcrowding of central city locations.
To summarise the discussion on determinants of location, the difference in nature of work and services produced by IT firms and other producer service firms, there are some basic differences in the factors influencing location. Unlike the findings from existing theory and empirical research, the stress is not on factors of communication and face to face versus indirect contact, but on aspects relating to intra firm communication and organisational factors. Since firms located in Bangalore have very little local market orientation, their inter firm contacts are more focused to their parent or subsidiary firms located elsewhere. The firms have more backward linkages with the city than forward linkages. Their primary interaction with the city is in terms of availability of skilled labour. In such a situation, it is organisational and other factors such as policy and value preferences that determine their location at an intrametropolitan level. In a very simplified version, firms locate in the city, expand and multiply offices, expand further and consolidate selected functions in peripheral locations.
Airoldi A., Janetti B.J., Gambardella A. and Senn L., 1997, "The Impact of Urban Structure on the Location of Producer Services", Service Industries Journal, Vol. 17, No. 1, pp. 91-114
Beyers W.B., 2000, "Cyber Space or Human Space", in Cities in the Telecommunication Age: Fracturing of Geographies, Wheeler, J.O., Yuko, Aoyama and Warf, Barney eds., pp. 161-180 (Routledge, London)
Bondenman E.J., 1998, "The Suburbanization of the Institutional Investment Advisory Industry: Metropolitan Philadelphia, 1983- 1993", Professional Geographer, Vol. 50, No. 1, pp112-126
Campbell H.S. and Harrington J.W., 1997, ";The Suburbanization of Producer Service Employment,"; Growth and Change, Vol. 28, pp. 345-370
Castells Manuel, 1989, "The Informational City" (Basil Blackwell, Oxford, UK)
Cervero Robert, 1989, " America's urban centers: The landuse transportation link" (Unwin Hymen, Boston)
Champion T, 2001, "Urbanisation, Suburbanisation, Counterurbanisation, and Reurbanisation", in Paddison, R. ed. Handbook of Urban Studies (Sage, London)
Coffey W.J., Drolet R. and Polese M., 1996, "The intrametropolitan location of high order services: patterns, factors and mobility in Montreal", Papers in Regional Science, Vol. 75, pp. 293-323
Coffey W.J.,2000, "The Geographies of Producer Services", Urban Geography, Vol. 21, pp. 172-183
Daniels P.W. ed., 1979, "Spatial Pattern of Office Growth and Location" ( John Wiley and Sons, New York)
Daniels P.W., 1985, " Service Industries: A geographical appraisal" ( Methuen, London.)
Dicken Peter, 1998, "Global Shift: Transforming the World Economy", Third Edition (Paul Chapman Publishing Ltd., New York)
Gad G.H.K., 1979, Face to Face Linkages and office Decentralisation Potentials: A study of Toronto,pp. 277 - 323, in Spatial Patterns of Office Growth and Location, Daniels ed. (John Wiley and Sons, New York)
Garreau Joel, 1991, " Edge City: Life on the New Frontier" ( Doubleday, New York)
Goddard J.B. and Morris D., 1976, "The communications factor in office decentralization", Progress in Planning, Vol 6, pp. 1-80.
Goddard J.B., 1975, "Office location in urban and regional development" (Oxford University Press, London)
Gottmann Jean, 1983, "The coming of the transactional city", Lectures delivered at the Institute of Urban Studies, ( University of Maryland, College Park, MD)
Hanna Nagy, Boyson Sandor, Gunaratne Shakuntala, 1996, "The East Asian miracle and information technology : strategic management of technological learning", World Bank discussion paper ; 326 ( World Bank, Washington D.C.)
Hartshorn T.A. and Muller P.O., 1986, "Suburban Business Centers: Employment implications, U.S. Department of Commerce, Economic Development Administration, Washington D.C" cited in Stanback,1991 and Champion ,2001
HDR, 2001, "Human Development Report" ( Oxford University Press, New York)
Henderson Jeffrey 1989, "The globalisation of High Technology Production: Society, space and semiconductors in the restructuring of the modern world" (Routledge, London)
Illeris Sven, 1996, "The Service Economy: A geographical approach" (John Wiley and Sons, UK)
Leslie Deborah, 1997, "Abandoning Madison Avenue: The Relocation of Advertising Services in New York City", Urban Geography, Vol. 18, No. 7, pp. 568 - 590.
Marshall J.L. and Woods Peter, 1995, "Services and Space: Key Aspects of Urban and Regional Development, Longman Scientific and Technical" (Co published with John Wiley and Sons Inc., New York)
Meier R.L., 1962, "A communications theory of urban growth" ( M.I.T. Press, Cambridge, Massachuesetts)
NASSCOM, 2002, "IT Industry in India: Strategic Overview 2002" (National Association of Software and Service Companies, New Delhi)
Niles J.M., Carlsonm R., Grey P.and Heineman G., 1976, "Telecommunications - Transportation Trade-Offs: Options for Tomorrow" (Wiley, New York)
Parthasarathy Balaji, 2000, "Globalisation and Agglomeration in Newly Industrialising Countries: The State and the Information Technology Industry in Bangalore, India", Doctoral Dissertation, University of California, Berkeley
Pye Roger, 1977, "Office Location: The Role of Communications and Technology" pp. 239 - 277, in Spatial Patterns of Office Growth and Location, Daniels ed. (John Wiley and Sons, New York)
Saxenian Anna Lee, 2000, "Bangalore: The Silicon Valley of Asia?", Paper presented for the Conference on Indian Economic Prospects, Center for Research on Economic Development and Policy Reform, Stanford, May, 2000
Schiller Russell, 2001, "The Dynamics of Property Location: Value and the factors which drive the location of shops, offices and other land uses" (Spon Press, London)
Stanback T.M.,1991, "The new suburbanisation: Challenge to the Central City" (West View Press, Boulder, CO)
Storper Michael and Venables Anthony J., 2002, Buzz: Face-to Face Contact and the Urban Economy, Unpublished Paper, http://econ.lse.ac.uk/staff/ ajv/research_material.html#buzz
STPI, 2004, "Information on Web Site of BangaloreIT.com", Department of Information Technology, Government of Karnataka, www. bangaloreit.com
Thorngren B., 1973, "How do contact systems effect regional development?", Environment and Planning, Vol. 2, pp. 409 -27
Zhou Yu, 1998, "Beyond Ethnic Enclaves: Location Strategies of Chinese Producer Service Firms in Los Angeles", Economic Geography, Vol. 74, No. 3, pp. 228 - 252.
Work Completed at: Dr.ingenior Programme, Department of Urban Design and Planning, Norwegian University of Science and Technology, N 7491, Trondheim, Norway
Edited and posted on the web on 22nd April 2005
Note: This Research Bulletin has been published in Environment and Planning A, 40 (2), (2008), 446-463