I would like to have read it but couldn't import it into Lotus Wordpro 96
or into MS Word 2.0, dunno why. If Anthony reposts it a s a straight text
file that could be good.
Mark
----------------------------------------------------------------------------
--------------
Note: Only text: all the graphs are missing (not transferable).
Technology Leapfrogging: The Software Challenge in India
Anthony P. D=Costa
Associate Professor
University of Washington, Tacoma
1900 Commerce Street, Tacoma WA 98402, USA
1. INTRODUCTION
During the last decade the Indian software industry expanded by leaps and
bounds. This growth was largely driven by export expansion. In 1987
software worth $54 million was exported from India (Heeks 1996). By
1997C98 fiscal year, the export figure stood roughly at $ 1,700 million
(Deccan Herald, June 20, 1998). Domestic software market is estimated to
be roughly one billion dollars and considerably more if in-house software
activity is taken into account. This is the only industry in India in
which the volume of export is larger than the domestic market. For the
Indian economy this is no ordinary feat and euphoria over the industry=s
performance is not all together unwarranted, given that the Indian economy
continues to be plagued by rampant poverty and persistent technological
backwardness.
The objective of this paper is not to explain this phenomenal growth but
rather to investigate whether any relationship between this massive
expansion of the software industry and innovative behavior exists and
whether the Indian industry is exhibiting technological leapfrogging.
Given the overtly export focus of the industry, I argue that the Indian
software industry represents innovativeness and leapfrogging only in a
limited sense. While export markets remain commercially lucrative for
India-based firms, the innovative capability is dictated by the needs of
foreign markets. The domestic market remains undeveloped and it fails to
push the industry to new technological heights. The externally-driven
software industry has thereby locked-in the Indian industry by truncating
any favorable externalities that might have existed with a dynamic domestic
software market. The absence of a strong producer-user dynamic implies
that the Indian industry, despite exhibiting one of the highest rates of
growth, does not necessarily display a penchant for innovations. This is
observable from the exports of low value software output. A policy that
strengthens the Indian software market can be made the basis for innovative
capacity.
The paper is divided into three main sections. The first section briefly
discusses the producer-user relationship in the context of innovation and
leapfrogging. The second section presents some of the salient elements of
the sector=s evolution in India, demonstrating the outward-orientation of
the industry and the inherent limits of the Indian industry. It critically
assesses the industry=s export performance to highlight the structural
obstacles associated with an externally-driven industrial evolution. It
also examines the extent to which exports have led to the technological
deepening. The third section presents the evolution of the industry in
terms of technological capability. Acknowledging some of the achievements
of the Indian industry, this section also outlines a national innovation
policy for incubating new areas of software development.
2. INNOVATION AND LEAPFROGGING: THE USERCPRODUCER SYNERGY
Innovative activity as we know it is generally integral to a dynamic
economy. The Indian economy until the 1980s was sluggish with
technological change limited to assimilation of foreign know-how.
Excessive regulation with entrenched institutional lethargy stifled
innovative activity. Since the 1980s gradual liberalization of the Indian
economy integrated the Indian software industry with the global one. The
worldwide rapid growth of the information technology sector provided a
window of opportunity for Indian workers. Foreign firms already accustomed
to sourcing raw materials and labor from distant, diverse markets, sought
to take advantage of the low-cost, high-skilled workers from India. With
software exports in rupee terms growing at nearly 50% per year it can be
safely assumed that there is some innovation in the industry. Sectorally
speaking the development of the software sector does represent
technological leapfrogging, given the largely agrarian structure of the
Indian economy, the technological backwardness of manufacturing industry
and the infrastructural sector, and a service sector dominated by
government employees and informal sector workers.
Leapfrogging, however, implies a set of stages, one or more of which can be
skipped. It also means that innovation is the driving force behind robust
economic growth. But the absence of a dynamic local economy constrain
innovative activity. Innovation is conditioned by a host of factors but
one important element is networking or information-sharing. For example,
in manufacturing a feedback loop either from the consumer or user of
intermediate inputs or capital goods is critical to information pooling for
innovation. Close proximity of firms, as in an industrial district or the
tightly knit hierarchical subcontracting system, foster firm flexibility
(see Robertson and Langlois 1995). In both systems, innovation rests on
information flow which is facilitated by proximity of buyers and sellers
and the institutional mechanism in co-designing and transferring technology
(D=Costa 1997). In other words, local interfacing reduces vulnerability
and encourages innovations. The question is what happens when the market
is international. Do we have the same kind of information flows that
encourages innovation? Are high exports a sign of leapfrogging?
Experience from East Asian export-driven economies shows that international
markets can foster quality manufacturing. Yet it is also clear that
innovation in East Asian casesCthat is the gradual movement up the
value-chain within and across sectorsCwas possible by strategic investment
and institutional support for reducing transaction costs and increasing
information flow (Dodgson 1997). An industry and technology policy was
integral to creating a strong domestic market and sustaining the learning
environment to move into high value production. Leapfrogging therefore
becomes possible when manufacturers interface tightly with users in a
closed loop fashion, sharing technical and marketing information and
cooperating financially in new ventures of mutual interest.
In a poor economy, such as India, limited demand and low quality
consciousness contribute to a systemic weakness in user feedback.
Cooperative institutional arrangements between buyers and sellers, which
might compensate for this, are weak as well. Software like capital goods
is embodied knowledge and it is the quintessential intermediate input, used
in an infinite number of ways in a variety of economic, commercial, and
intellectual Adomain@ settings. A domestic base for software production
acts as the foundation for learning by doing, using, and interacting (see
Niosi 1991). The mastery of these integrated processes leads to innovative
responses by firms operating in a highly competitive and volatile market.
In the absence of robust producerCuser links in the Indian economy,
self-evident from the technological and organizational weaknesses of
manufacturing, banking, and scientific effort, the Indian software industry
can be hypothesized to be non-innovative. The development of the Indian
software industry is likely to be hobbled by the absence of a dynamic
domestic user industry. This contrasts with the highly differentiated US
economy which is both a heavy user, producer, and exporter of software.
Among other things, greater the density of software usage in the domestic
market (the missing component in India=s vast market) the tighter will be
the producerCuser linkages.
Perhaps the most important argument for a tight producer-user network is
technological convergence (see Athreye 1998). In the software industry
innovations when diffused widely across economic sectors and within
manufacturing can be a solid foundation for innovation, especially when
directed to solving specific local problems. With a dynamic domestic
sector it becomes realistic not only to innovate (the market pull factor)
but also Aapply@ these innovations in a specialized manner (the supply push
factor) in a wide range of commercial activities. A large user market
supports the diffusion of software applications, making innovations
commercially viable. With the spread of computerization in the local
economy the demand for software is likely to increase significantly. More
importantly as the needs themselves diversify, software applications
(custom-made or packaged) are also expected to embody complex solutions in
a re-usable manner. The resulting synergy becomes the foundation for
innovative software development.
In theory, the argument for producerCuser networks does not distinguish
between domestic and foreign firms. In other words, foreign users could
very well support domestic Indian software industry. While this is
plausible the absence of a strong local sector introduces structural
barriers to innovative capacity. For example, because of weak domain
expertise, such as ATMs and embedded software in manufacturing, the Indian
industry is likely to specialize in areas that do not demand such
knowledge. The technical spin-offs that emerge from external demand does
not substantially contribute to the producerCuser circuit. The reasons are
not difficult to identify. First, high export revenues locks the industry
into low-value, albeit lucrative segment of the software industry. This is
typically the on-site services provided by Indian engineers overseas.
Second, as costs of doing on-site work increases due to higher
(dollar-denominated) salaries and attendant political pressure to reduce
foreign engineers, on-site projects become less preferred to off-shore
development. In both cases the foreign partner has the final say in
project management and the content of work. In Aon-site@ work the client
virtually determines all project requirementsCfrom manpower needs,
technical specifications of the projects, to detailed project management.
Thus the industry=s less glamorous Abody shopping@ phrase is not off the
mark.
The shift from on-site to off-shore could theoretically translate into a
movement from low value to high value work, especially if R&D is part of
the off-shore work. However, moving up the value chain itself becomes
problematic if the industry=s dynamic is not localized. First, off-shore
work is for the parent firm (and with some exceptions) is simply services
from off-shore sites. They do not necessarily lead to more complex tasks.
Data links facilitate work to be conducted off-shore. India=s geographical
location permits access to idle US computers. Second, because off-shore
work is for a market internal to the firm, the revenues generated is not
value based. Instead subsidiaries of MNCs are cost centers, reimbursing
the off-shore site for costs incurred and not remunerating for value
created. Here lies the structural strength of MNCs in tapping and managing
foreign technical competence. Third, whatever R&D is done off-shore
remains within the firm, typically headquartered overseas. While this is a
requirement for competition in imperfect markets, the intellectual property
right remains with the firm. The off-shore site does not generate a flow
of revenues which could otherwise have been secured with licensing of
innovations. Finally, the reliance on foreign markets (through exports of
services) discourages the industry in undertaking packaged products.
Shrink-wrapped packages are the most lucrative as they are re-usable. They
also demand solid financial backing for both R&D and marketing purposes,
none of which are Indian strengths.
>From the Indian point of view an externally determined feedback loop can
sustain a dynamic industry as long as growth and diversity of the software
market continues. But as we have argued, foreign demand for services is
likely to be in select areas. All said and done, the continued development
of the industry cannot solely rely on exports. A vibrant domestic user
industry is essential for the growth and development of this sector. In
the absence of synergy, feedback mechanisms are likely to be weak, thus
retarding the development of Adeep knowledge@ critical to the industry. It
is obvious that a technology policy, whether an economy-wide or more
sectorally-focused approach, is warranted. Software development is highly
skill-intensive, requiring substantial investment in basic technical
training, which the state has thus far provided. However, only a wider and
deeper exposure to the entire gamut of Adomains@ can allow the Indian
software sector to flourish. These domains need to be established at home,
promoting not only technological convergence in a generic sense but also
raising the overall productivity of the economy in a developmental sense.
3. CHALLENGES TO THE INDIAN SOFTWARE INDUSTRY
3.1 The Development of the Indian Software Industry
Between 1980 and 1996 exports expanded from $4 million to over $1 billionCa
nearly 250 fold growth (Figure 1) (NASSCOM 1998). The compounded annual
growth rate between 1992-93 and 1996-97 was over 48%. However, real growth
rates reflect nearly half as much (Sen 1994: 56). Between 1987-93, nominal
growth was 46.4%, while the real growth was 28% per annum. As shown by Sen
the devaluation of the Indian rupee vis-a-vis the US dollar has contributed
significantly to the export trend. Nearly 60% of the exports was directed
to the US market. While this is a sizeable market share from the Indian
side, the share of Indian exports to total US market was less than 2%.
In disaggregating the export data we find that the bulk of exports (58.7%)
comprised on-site services, while less than a third was off-shore services
(Table 1). The remaining 11% was grouped under off-shore packages.
On-sites services generally comprises low value workCtypically coding and
testing. The availability of low cost manpower with good programming
skills has been the basis for high growth of exports. In breaking down
software development by skill category we find that very little of
conceptual or designing aspects of software have been behind India=s export
thrust (Heeks 1996). Indian workers do not undertake design and are not
exposed to challenging project management assignments. Instead they follow
the design specifications established by the foreign software clients. They
do not require Acreativity, organizational understanding, or consultation
with end users (Heeks 1996: 82, emphasis added). Hence, the growth in
exports can be seen as a linear expansion of manpower without significant
deepening of skills. As the demand for on-site services increased, Indian
firms aggressively sought to create a niche for themselves in this market
segment. A major shortcoming of on-site services is the non-reusability of
the software created by the service provider. Only in a few cases the
experience gained from client-specific work was transformed into software
products. This means that unlike packaged software there is only one-time
flow of revenue, limiting the opportunities to maintain a momentum of
investment in high-skill workers to carry out R&D.
Table 1
Types of Software Exports
(US $ million)* Percentage
On-site Services 572.25 58.7
Off-Shore Services 294.50 30.2
Off-shore Packages 108.25 11.1
Total 975.00 100.0
Source: NASSCOM (1998).
* at Rs.40/US$
Another problem of service-oriented exports has been the huge absorption of
Indian manpower in the AYear 2000 (Y2K)@ problem, plaguing older computer
systems in the industrialized countries. The irony is that the Y2K
problem, although commercially worth several hundred billion dollars, by
itself is not a frontier moving problem. It requires a large pool of
workers, tediously identifying date fields, fixing them, and testing the
fixes. In the short-term this is a highly lucrative activity and hence the
Indian software firms are content to supply an increasing headcount of
fairly low-level, homogeneous, skilled programmers for finding the Y2K
solution in the export markets. Critics of this approach have pointed out
a massive misallocation of highly skilled engineers workers away from
potentially high value-added activities to more tedious code writing. The
pull of the Y2K export services market has been strong enough to thwart any
strategic vision or a planned response to the glut of workers that might
arise after the Y2K problem is taken care of. While similar services might
be needed even beyond the year 2000 and for the gigantic conversion
problems associated with European integration, it is not clear if such
efforts will generate innovative knowledge elevating the Indian industry
onto a higher technological plane.
3.2 Exports and Technological Deepening?
It has been reported that on-site services are increasingly shifting toward
off-shore services. This has been also suggested as the wave of the future
and a business strategy (see Figure 2). Off-shore services have been made
possible with high-speed data communication links, which allow Indian firms
to transfer software to their foreign clients from India itself. The
nature of work itself has not necessarily undergone fundamental
transformation. Work remains similar to on-site services. However, a
certain amount of managerial capability is transferred as local managers
oversee several projects. There is also a rising trend to create packaged
software under off-shore services.
With the expansion of off-shore projects, additional skills are generated
through systems integration, managing facilities, and working on software
conversions for multiple hardware platforms. These are certainly
value-adding activities but they still comprise a small proportion of
off-shore work. Much of the services continue to be client-specific and
client determinedCeither overseas client or a MNC subsidiary at home. A
large chunk of the off-shore work also includes data entry, the lowest of
the skill category. From an employment point of view this is a favorable
development but does not contribute much to the innovative potential.
Whatever off-shore packaged product development takes place in the country
are MNC brands. There is no intellectual property right retained in India
or by Indians, even if they might have contributed to the development of
such branded products. Similarly, domestic markets are unable to compete
with MNCs since they do not have the same kind of market clout. The
strategy of MNCs is understandable. They seek cheap, skilled labor,
remunerate them well, and reap the benefits of innovations for their
commercial success. The inability of the Indian industry to retain
re-usable software knowledge is largely a structural and strategic problem.
The lure of dollar export revenues on the one hand and the
Ainternalization@ of innovation within the MNC but across national
boundaries cuts of the Indian industry from undertaking any significant
self-initiated innovative ventures. The problem is compounded when both
on-site and off-shore work exacerbates an outflow of skilled workers
overseas. Because of the weak link between domestic and export software
markets, skill transfers to the former from the export market is
negligible. Instead job hoppers typically move within the software sector
or else go overseas.
Another disconcerting factor is the lack of systems approach by Indian
firms to quality and productivity enhancement. Very few firms maintain and
monitor any productivity-related data systematically. More importantly, if
software engineers are engaged in code work, their underemployment is
likely to lead to lower productivity. In the early 1990s, Indian
productivity was considered to be about 30% of the US (Sen 1994). With
limited project management experience and unstructured approach to software
development, it is not surprising to find that very few firms have systems
of procedures in place. It is only now that firms are beginning to seek
the ISO 9001 (European standard), IS 14000 (Indian standard) , and IEEE (US
standard). In 1997-98 there were nearly 60 firms with ISO 9001/2
certification, that is about 13% of the members of the National Association
of Software and Service Companies (NASSCOM). NASSCOM=s members contribute
nearly 95% of Indian software revenues. Certification by itself is not a
reflection of innovativeness. It is the documentation of a set of
standardized procedures that have been introduced in the firm=s operations
to assure quality. Of course systematic procedures is the one of the first
steps to increasing productivity and demonstrating a proclivity toward
innovativeness.
There is another industry standard set by the Software Engineering
Institute (SEI) at Carnegie Mellon University in Pittsburgh, USA called the
Capability Maturity Model. It is an evolutionary model of firms moving
from Aan ad hoc, immature process to mature, disciplined process@ (NASSCOM
1998: 63). There are five levels of maturity with Motorola=s Indian
operations having attained a level 5, a feat that has been matched so far
only by IBM=s NASA project. There were six other firms that attained
levels 3 and 4. However, all except one firm was a multinational
subsidiary, indicating the structural difficulty in wresting control from
MNCs. On the other hand, Motorola=s achievement was downplayed by its
managing director as fairly routine. WIPRO a wholly-owned Indian firm has
instituted General Electric=s ASix Sigma@ quality standard, which at level
six translates to only 3.4 defects per million and at level five 230
defects per million. It is clear that to be innovative the least that a
firm could do is establish a set of standard procedures, document them, and
periodically upgrade them to ensure quality and productivity.
Software is a high technology, high-skill activity. The Indian software
industry has achieved international recognition. However, on closer
inspection there are significant challenges for the industry (Figure 3). A
weak domestic sector has resulted in an undue reliance on foreign markets
creating a lock-in effect in low value work. Additionally, export of
services is increasingly undermining India=s cost competitiveness by
mopping up the limited pool of skilled workers, salary inflation, and a
high labor turnover of 30%.
4. INCUBATING INNOVATIONS IN THE SOFTWARE INDUSTRY
4.1 Emerging Technological Foundation
The Indian software industry despite exhibiting remarkable growth faces
serious structural problems, especially in building long-term innovative
capacity. The main reason has been an unusual reliance on the export
market and with successive linear expansion of on-site services in overseas
markets, the Indian industry has been locked-in into a niche that is not
high on the value chain. However, with off-shore services and packaged
product development, the Indian industry is gaining a foothold in the more
difficult markets. There have been some recent successes. For example,
Indian engineers at the Texas Instruments= Bangalore center designed a
digital signal processor (DSP) which could be used in a wide range of
consumer durables, digital cameras, robotics and other industrial
applications. Ncore an information technology design services company has
licensed its modem software and speech compression technology to
Mitsubishi, LG Semicon, and a couple of US firms. Similarly WIPRO sold an
IPR (a synthesizer core) to a reputed Japanese computer peripheral
manufacturing company (Figure 4). The assignment included interfacing
WIPRO=s logics with the Japanese client=s logic, a jump from routine
programming to staying ahead of industry standards, such as the IEEE.
These examples reflect the creeping innovative capability of India-based
software firms. They also serve to illustrate the limits the Indian
software industry faces. Texas Instruments will hold the DSP patent and
only a portion of the stream of income accruing to it is likely to be
reinvested in India. Also the application of the DSP is for industries
which are not very strong in India. If, however, the domestic market were
to be strong the multiplier effect associated with a producer-user nexus
would have been larger. In the case of WIPRO, the technology was not
licensed. Instead, the company accepted a one-time fee under the Afee
bearing application-specific worldwide usage right.@ It is unclear why a
royalty based licensing arrangement could not be negotiated by WIPRO. That
WIPRO does not have the hard bargaining clout with Japanese MNCs at this
time cannot be ruled out.
There are several other examples of off-shore activity in India moving away
from pure client-services to more of R&D for new products. For example,
Siemens recently set up a center in India to produce integrated circuits,
CAD software, and cell libraries to serve consumer and automotive
electronics, and computer peripherals. Similarly Oracle will be investing
$40 million to work on oil and gas solutions, with its Indian operations
re-engineering the core solution obtained from British Petroleum. Cadence
Design Systems (India) is supporting its electronic and system design
automation software. It plans to create a full fledged customer service
response center for South East Asia. Other products with business
applications are in the offing. Indian firms have also begun using
state-of-the-art technologies, such as client-network computing, intranet,
computer aided software engineering (CASE), fourth generation language
(4GL) tools, graphical user interfaces (GUI), and open systems (Elsoftex
1997). These developments support the notion that Indian software industry
is doing value-added work and by extension creating an innovative capacity
for the future. However, as has been shown the pull of the export market
has been much too strong for most small and medium firms to engage in
low-level services. Only a handful of large firms, mainly subsidiaries of
MNCs and a few domestic firms, are engaged in some innovative work. These
are firms that are all export-driven and hence the externalities associated
with innovations are not captured within the national economic system. The
problem is the weak links to the rest of the industrial and service
sectors.
4.2 Toward a National Innovation System
>From the preceding analysis it is evident that the basic thrust of the
Indian software industry has been toward exports. This is not surprising,
given the weak domestic market and lucrative foreign markets. It also fits
with the current economic ideology and policy context. Free market and
export-orientation are argued to increase national welfare and countries
wishing to improve their lot should pursue such policies. It is thus
fitting that the World Bank sponsored a study to examine the export
potential of the Indian software industry (1992). As such exports appear
to be superior to not exporting. However, as has been shown, in India the
exports have been at the expense of the domestic market. This has worked
through the huge demand for Indian software personnel to service foreign
clients. The Indian industry as a result has been locked-in into low value
software production. The problem is inherently structural and hence little
can be done with policy intervention in the immediate future.
Nevertheless, it is clear that a long-term innovation policy is crucial for
the Indian industry to move up the value-chain and prevent outpricing
itself in the world market.
Several suggestions have been madeCby the industry, government, and
practitioners. These cover infrastructure development, education and
training, and fiscal incentives. India=s physical and communication
infrastructure is far behind international standards. Some effort have
been made in this direction. Telephone lines, satellite communications,
and dedicated data links have been installed. The Software Technology
Parks of India under the Department of Electronics has instituted wireless
data communication systems for export-oriented firms. However,
international and intranational connectivity still remains underdeveloped.
The Indian technical education is inadequate to meet the growing demand
for skilled workers. The engineering institutes do not necessarily provide
the kind of exposure necessary to undertake data link-based software
development nor is it the case that engineers need to do coding work. As a
policy measure not only should curriculum be modified to introduce new
areas of software development but also to train non-engineers
state-of-the-art programming skills. Patterned after the successful Indian
Institute of Technology, institutes of information technology have been
recommended. Various fiscal incentives are offered to the software
industry. For example, import duty for software and hardware for exports
is zero.
These steps in support of the software industry are in the right direction.
But they are piecemeal and do not amount to a systematic innovation policy
based on institutional arrangements. The recent draft National Informatics
Policy tries to integrate some of these elements with that of institutional
mechanisms for diffusing computerization broadly. But there appears to be
little emphasis on producer-user links. It is acknowledged that
information technology density in India is very low and that
computerization should be increased. Lowering prices of both hardware and
software would be one way to increase computerization. Prices can be
lowered if there is scale economies. Specialization is one way to attain
high volume production. However, as Sen (1994) points out the Indian
industry is fragmented with very few specializing in their core
competencies. A more focused approach on the part of Indian firms is also
likely to deepen their knowledge base. Incentives designed to increase R&D
would be a major determinant for future innovations. There are at least
three ways to do it. One way would be to provide fiscal incentives for
both domestic and foreign firms for their R&D effort. Promotion of venture
capital will be critical to the industry=s innovative capability. The
second would be to create a link between academia and private business. As
of now this link is virtually absent in India. In the software sector,
business-academia partnership is easier to establish due to reduced capital
costs. The third institutional mechanism is to tap the vast expertise
which exists in various government departments but which are hidden from
public view because of their strategic or military orientation. These
include the Department of Space, Department of Telecommunications, and
their joint venture under the Indian National Satellite System, the Indian
Space Research Organization, among others. Software projects such as
geostationary positioning systems used in satellite applications can be
applied in the civilian sector for navigation purposes.
There are other frontier-moving applications of software. They include
areas of biotechnology, fuzzy logic, neural nets, genetic computing, and
VLSI chip design, among others. Relatedly, the Indian software industry
has not tapped the huge market for embedded software. The embedded
software is found virtually in all manufactured products, such as toys,
tools, appliances, and electronic gadgets. The East Asian countries, such
as Japan, Korea, and Taiwan provide a massive opportunity in embedded
software, a market that has not been tapped by Indian software firms. As
the Indian economy moves up to a higher growth trend, the inevitable
deepening of industrialization will create a massive user industry. Market
research and trade and industry liberalization with a systematic
technology policy ought to be a part of the overall strategy in developing
this segment of the Indian software industry. Government funding of
research in the software industry will be necessary as well. Currently
there is no such program in place.
5. CONCLUSION
In this paper I made a case that the Indian software industry despite rapid
export growth still belongs to low value on-site services. Foreign
exchange earnings and relatively good remuneration notwithstanding, there
are several structural weaknesses of the Indian industry. Its aggressive
stance toward exports, promoted by the industry, the government, and by the
firms themselves, has locked-in the industry into a Alow road to
accumulation.@ The high level technical skills deployed for such services
have a high opportunity cost. They would be better placed in carrying out
more complex design and project management tasks. Some firms have begun
doing this on a small scale but more needs to be done. To capture the
benefits of the export sector for the domestic segment an innovation policy
was suggested. A balanced development of the industry is critical for
maintaining and upgrading its competitiveness. Investment in communication
infrastructure and in education and training combined with institutional
mechanisms to promote R&D are obvious areas of policy intervention. There
are also other areas, such as frontier-type applications of software and
greater market presence, such as in the embedded software market of East
Asia. These steps are an integral part of an innovation policy that aims
to support a large, dynamic domestic market and which will be expected to
have a synergistic relationship to the export sector. The innovation
policy presented here is skeletal. Additional research is needed to
understand how the Indian industry is evolving and how it can be made to
interface with its partners in industry, government, and academia. An
innovation policy must push for the rapid diffusion of information
technology in the domestic market. It must encourage re-usable software
product development and implement large-scale projects, such as Computer
Maintenance Corporation=s mammoth railway reservation system, to boost
solid project management skills. In the absence of such measures the
Indian software industry will find it difficult to extricate itself from
the groove of on-site services and experience leapfrogging only
momentarily.
REFERENCES
Athreye, S. 1998, On Markets and Knowledge, ECRC Centre for Business
Research, Working Paper 83, University of Cambridge.
D=Costa, A.P. 1997, Economic Governance in the Indian Automotive Industry:
Flexible Institutions for Mass Production Goals, unpublished paper.
Deccan Herald 1998, June 20, Bangalore.
Dodgson, M. 1997, Innovation Policies in East Asia, 1st International
Conference on Technology Policy and Innovation, Macau, July 2-4, 1997.
Elsoftex 1997, Elsoftex Newsletter, August.
Heeks, R. 1996, India=s Software Industry: State Policy, Liberalisation and
Industrial Development, Sage, New Delhi.
International Data Corporation (IDC), Company Documents.
National Association of Software and Services Companies (NASSCOM) 1998, The
Software Industry in India: A Strategic Review 1997-98, New Delhi.
Niosi, J. (Ed.) 1991, Technology and National Competitiveness: Oligopoly,
Technological Innovation, and International Competition, McGill-Queen=s
University Press, Montreal.
Robertson, P..L. & Langlois, R.N. 1995, Innovation, Networks, and Vertical
Integration, Research Policy, vol. 24, no. 4, pp. 543-562.
Sen, P. 1994, Software Exports from India: A Systemic Analysis, Electronics
Information &Planning, November, pp. 55-63.
World Bank 1992, India=s Software and Services Export Potential and
Strategies, IDC (India), New Delhi.
ACKNOWLEDGEMENT
The research for this paper was supported by a Senior Fellowship from the
American Institute of Indian Studies. Conference attendance is made
possible by the South Asia Program, University of Washington, Seattle and
the University of Washington, Tacoma. I would like to thank NASSCOM, STPI,
Delhi and Bangalore, and the numerous software firms in Bangalore, New
Delhi, NOIDA, and Calcutta for sharing their insights on the software
industry.
