Speech to World Economic Forum in Davos, 2000
The King actively promotes the Jordanian software export sector
Software exports have become a cause for excitement in dozens of developing nations.
These are nations that are all searching for the recipe to become the ÔÇťnext India.ÔÇŁ India,
once known as a land of poverty, has now become an IT superpower. Major Western
software companies, including Microsoft, Oracle, and SAP, are using Indian software
centers to develop their products. IndiaÔÇ™s independent ´¬ürms compete with the top ´¬ürms
from the industrialized nations.
This excitement is understandable since many developing countries are primarily
exporters of commodities, such as cotton or coffee. Other export sectors, such as tex-
tiles, handicrafts, assembly, or manufacturing, offer primarily low-skilled employment
with low wages, often with ÔÇťsweatshopÔÇŁ working conditions. And then in the 1990s,
along comes software, a high-skilled industry suddenly within reach of developing
nations. The current economies of software production, in particular its low-capital and
high-labor intensity, is especially attractive for low-wage, labor-surplus economies.
Another allure is that the offshoring industry continues to grow year after year and
that demand is forecast to continue rising for some years to come.1 Nearly all observers
agree that we are far from the saturation point. This growth creates new business
opportunities for developing nations to attract employment and income-creating work.
Governments, policy-makers, academics, and journalists are infected by this exciting
potential. International organizations, such as the World Bank and UNCTAD, have
recognized the economic and social impacts of a software export sector. National aid
200 Other stakeholders
agencies from the US, Japan, Germany, Switzerland, and The Netherlands, among
others, have all become interested in helping developing nations move forward by
growing their software export industries.
The production of software is relatively environment friendly and does not depend
on roads and harbors. Information, unlike products such as textiles or automobiles, can
be transported quickly and cheaply through digital channels. Moreover, the required
investments are modest, particularly when compared with industries, such as machine
building or steel plants. Mastek, now a major Indian software company, began in a
garage with only a few programmers and one fax.
There are about 150 developing nations in the world. This is a heterogeneous group,
consisting of both low- and middle-income countries. Many developing countries have
already begun software exports. These countries include some of the poorest nations in
the world, such as Bangladesh and Nepal. Other nations are making preparations to enter
In Chapter 4 (ÔÇťThe Offshore Country MenuÔÇŁ), software-exporting nations were clas-
si´¬üed into three tiers. Based on the criteria of industry maturity, export revenues, and
clustering, India and China are Tier-1 nations, along with many industrialized countries.
They have large numbers of IT enterprises and their huge software exports are growing
fast. Tier-2 nations are the emerging exporters, such as the Philippines, Malaysia,
Pakistan, Mexico, or Brazil. They already have signi´¬ücant export volumes, although the
number of exporters is usually less than 100. Tier-3 countries are still at an infant stage
of exporting; examples are Cuba, Iran, Vietnam, and Indonesia. Their number of export-
ing ´¬ürms is small, and so their volume of export work. This chapter is targeted at
policy-makers, government of´¬ücials, and other stakeholders from the Tier-2 and -3
Choosing a national strategy
Once exporting software has caught their attention, nations in Tiers-2 and -3 are faced
with a number of policy choices. Figure 10.1 depicts a simpli´¬üed policy decision tree.
While, it is unlikely that national policy-makers travel down this decision tree in a
methodical fashion, it is a useful model to introduce the policy choices and trade-offs
inherent in these choices.
Policy Decision No. 1 questions national investments in IT. Given the limited resources,
is it worthwhile to focus those resources on the production sector (the sector that
produces software)? Or, rather, should the country focus its resources on using IT for
broader information society objectives, such as universal access, broadband access,
e-government, or buying computers for schools? The debate about IT investments is
most intense when it pertains to the poorest nations: Should the nation invest in a
villageÔÇ™s computing and connectivity when the village itself lacks drinking water,
201 Building software industries in developing nations
Generic IT services
get Attract foreign R&D
t /t a
ec No. 3
Sel us Software products
ctio Policy Decision
p ro No. 2
Policy Decision stru
No. 1 e and
ena ÔÇó Human capital
s ÔÇó Technology parks
Figure 10.1 Policy decision tree for choosing a national strategy.
paved roads, and electrical power? As demonstrated in the Indian state of Andhra
Pradesh and in research by the World Bank,2 there are tangible results in using IT for
reaching the poorest sectors of society by, among other, improving public administra-
tion, delivering remote health care services, distance education, and achieving gender
equality. Dozens of developing countries have formulated national policies in these
areas, labeled ICT (Information and Communication Technology) policies.
We believe that Policy Decision No. 1 is not exclusionary. Investment in IT produc-
tion (software in this case) is needed to support broader information society objectives,
otherwise the poorest nations will rely solely on foreign aid and foreign specialists to
provide computing needs. Furthermore, as we will discuss in this chapterÔÇ™s next sec-
tion, there is evidence that nations bene´¬üt broadly from a strong software sector, par-
ticularly if it is an exporting sector.
Policy Decision No. 2 is perhaps the more controversial question: How to encourage
the software production sector? Should the software industry be encouraged by select-
ing a target vertical industry to focus national attention; or rather should the nation
invest in foundational, infrastructure, and enabling elements? The ´¬ürst of these choices
means that the nation, through its ministries and other governmental bodies, moves up
the tree to Policy Decision No. 3.
It is our belief that governments should not select vertical industries. Governments
tend to have a high failure rate when it comes to choosing on which technology to
place a bet. Markets and technologies tend to move faster than government policy-
makers. The best strategy for nations is not to choose a speci´¬üc software industry strat-
egy at all. The local companies will gravitate towards the more successful niches much
faster than government policy-makers are able to react.
Rather, governments should focus on those areas in which they have clear strengths
relative to market forces. These are areas that the market by itself cannot usually act
202 Other stakeholders
successfully. These areas tend to be in infrastructure and other foundational areas.
In particular, for software exports, there are two areas which have the greatest impact:
human capital and technology parks. Investment in human capital will make the nationÔÇ™s
industry more ´¬‚exible and competitive in the long run. After all, competition in software
is not just offering cheap labor; it is about low-cost skilled labor. Human capital invest-
ment includes programs in science and technology at national universities, as well IT-
related programs in specialized technical schools and vocational schools.
The second policy, which we believe has the greatest impact, is to facilitate the cre-
ation of technology parks. Such parks make it easy for local or foreign ´¬ürms to set up
operations. A technology park is made up of one or more buildings in which there is a
full range of services including reliable electrical power, high-speed lines, cabling,
physical security, and so on. Foreign technology ´¬ürms seeking new locations are espe-
cially drawn to these parks because they perceive them as being business friendly. And
many nations also give tax incentives to those companies, domestic or foreign, operat-
ing in these technology parks.
These two foci, human capital and technology parks, are not the sole areas for gov-
ernments to act. We note three other important areas that governments, together with
industry associations, can affect change:
Ô—Ć In the short term, provide marketing assistance for those ´¬ürms already exporting.
As we describe in Chapter 11, building relationships with potential foreign clients
require considerable efforts, especially if the provider is located thousands of
kilometers away. Governments can offer marketing assistance by funding trade
shows, leading delegations, and setting up match-making projects (to facilitate
´¬ünding clients in key target markets).
Ô—Ć In the longer term, promote collaborative efforts between the government, software
companies, ´¬ünancial institutions, universities, other educational programs, and
business associations. Today these are often labeled publicÔÇ“private partnerships.
Ô—Ć Assist ´¬ürms in attaining internationally recognized standards of quality. Such
standards address a fundamental international marketing problem faced by
providers from developing nations, namely, that clients see foreign suppliers as
exotic and risky. The government ministry, together with the industry association,
can set up training programs focused on quality certi´¬ücations.
Which focus to select?
Nations that do travel to Policy Decision No. 3 have another dif´¬ücult choice to make:
deciding on a focus for their software export industry within a highly competitive envi-
ronment. Some larger nations may choose to target all four foci depicted in Figure 10.1.
Choosing all four may dilute any advantage of focus. At the same time, nations can
choose more than one and there is some overlap between these foci.
203 Building software industries in developing nations
Target No. 1: Generic IT services
This is the default choice: to be a commodity low-cost service provider. This choice
does not appear to be glamorous since most Tier-2 and -3 countries compete primarily
on low wages. Dozens of national industries, including thousands of ´¬ürms, are offering
commodity skills in programming (e.g. using Microsoft C or Java) with little
national specialization and differentiation. Yet, just as there is room at the local fruit
and vegetable market for many undifferentiated merchants selling very similar toma-
toes side by side, there is room for a large global market in software programming
services. This is not necessarily an incoherent path for small countries to select since
most of them have no possibility of becoming the next Indian powerhouse. In fact,
Indian ´¬ürms face increasing competition from cheaper countries. By offering com-
modity skills, Tier-2 and -3 countries can ´¬ünd clients, build domain knowledge and
later, perhaps specialize in service niches.
Ideally, those nations that focus on providing software services should ´¬ünd an identi-
´¬üable niche. In such a case the nation builds a cluster of successful ´¬ürms exporting serv-
ices in an identi´¬üable specialty niche. This is desirable since differentiated services are
more pro´¬ütable than those that compete solely on price. Furthermore, differentiated
services reduce costs of marketing. While this is desirable, it is not easily attainable: we
are not aware of such identi´¬üable clusters in Tier-2 or -3 nations. Therefore, targeting a
services niche for a national industry, while possible, is a formidable task.
Clusters of specialty may emerge when they are rooted in some other national com-
petency. The model that policy-makers should strive for is to situate the industry close
to strong domestic technology users in order to create synergies with these users.3 For
example, Costa Rica has chosen as its national strategy to build its ÔÇťgreenÔÇŁ industries,
to capitalize on its well-recognized environmental assets and knowledge. Therefore,
software companies could situate themselves ÔÇťin proximityÔÇŁ to these green ´¬ürms and
develop expertise in their special needs.
Target No. 2: Attract foreign R&D activities
The nation can attract foreign technology companies which will set up wholly-owned
software development centers locally. This is quite desirable because the foreign ´¬ürms
provide capital and leading edge know-how. Additionally, foreign research and develop-
ment (R&D) centers are the locus of highly skilled innovative activities in most nations.
Most innovation activities in the new offshore destinations, including India and China,
still take place within subsidiaries of foreign technology ´¬ürms. For example, over the 25-
year period of 1978ÔÇ“2003, foreign-owned software ´¬ürms in India were authors of a total
of 110,914 copyrights (the majority from IBM), whereas Indian software ´¬ürms only had
208 copyrights.4 Attracting foreign R&D is a desirable but dif´¬ücult focus in which only
a few countries have had any success: India, China, Israel, and a handful of others. Many
developing nations are unlikely to be successful in this policy choice.
204 Other stakeholders
Target No. 3: Software products
Software products are attractive because, when successful, they can be far more prof-
itable than services. Furthermore, software products are launched as a result of greater
innovation intensity, which is a characteristic that nations strive to achieve. Unfortunately,
national success in exporting software products is quite dif´¬ücult. The basis for com-
peting in software products is deep domain knowledge (in business, science, or some
other domain) coupled with strong managerial and marketing skills. Most of the
nations that began competing in software lack these skills inside their industry.
Furthermore, software packages require capital, since the production investment
comes upfront. International marketing of software products is usually more dif´¬ücult
and costly than the marketing of services. Low labor cost is less important.
With the exception of Israel and Ireland, none of the new offshore nations has yet
made a mark in this sector. Even India has had very limited success, and that, only after
decades of intensive knowledge transfer from abroad;5 and after years of urging by
observers to diversify from low-end services. There are a handful of examples of single
´¬ürms from developing nations successfully exporting software products. Signum from
Ecuador succeeded in selling its Spanish language checker to Microsoft which incorpo-
rated the product in its Of´¬üce suite. STA, a ´¬ürm in the Philippines, developed and sold a
Year 2000 conversion tool abroad. Such examples, however, do not represent more than
the success of a single, individual software company, not of a national industry. We have
also found that some companies from developing nations attempt to market software
products that are inappropriate for the markets in industrialized nations. Their naivet├©
about client needs is another reason to stay away from software products.
Target No. 4: IT-enabled services
Smaller developing nations may even choose to migrate away from software or ÔÇťskipÔÇŁ
IT altogether and concentrate their national resources in specialized areas within the
IT-enabled services (ITES) sector. An example is the Philippines, which has seen suc-
cess in exporting IT services, but has been far more successful in exporting IT-enabled
Sri Lanka may also be shifting its attention.6 A window of relative peace from its civil
war created a new focus on opportunities for economic growth and exporting, and, in
particular, exporting IT-enabled services. The IT-enabled service ´¬ürms in Colombo have
traditionally focused on a few simple operations, such as data entry, but there is interest
in expanding into new areas at both the lower and higher end, including accounting,
legal, and engineering services. Engineering design drafting of the Dubai airport was
undertaken in Colombo. The ´¬ürst offshore call center was commissioned in 2002 and
uses voice over IP.7 Indian IT-enabled service providers saw opportunities in Sri Lanka:
for example, WNS Global Services began call center and transaction-processing opera-
tions. In the meantime, in 2002, the government embarked on an initiative to formulate
a comprehensive national ICT development strategy called ÔÇťe-Sri Lanka: ICT develop-
ment road map.ÔÇŁ In addition to the usual national ICT components, such as connectivity,
205 Building software industries in developing nations
legal reforms, and human resource (HR) development, the plan included some actions
aimed at the IT-enabled services export industry.
ITES: the policy choice of skipping software altogether
For some years now, the offshoring of IT-enabled services has grown faster than the
growth of IT offshoring. It is ÔÇťthe next big thingÔÇŁ in offshoring and many Tier-2 and -3
countries have focused their attention on this area. As we described in Chapter 1, off-
shoring ITES is closely tied to offshoring of IT in a number of ways: the providers are
often identical, the clients are often the same, the facilities are often comparable, and
the managerial know-how is somewhat similar.
ITES also has some important advantages vis-├á-vis IT: some IT-enabled service
areas have low-entry barriers in terms of skills, scale, technology, managerial capabil-
ities, or domain knowledge. This makes IT-enabled services more accessible to the least
developed nations. It is a sector in which even some African countries are active. For
example, Ghana, Mauritius, Morocco, Senegal, and Tunisia are hosting call centers,
some of which are French speaking and are targeting France.
Nations considering IT-enabled services have many specialties to choose from. These
specialties fall into one of the following three categories: customer interaction serv-
ices, back-of´¬üce operation services, and data and content integration.
Customer interaction services
These are front-of´¬üce activities, where employees are in direct contact with customers.
These services are mostly voice based, with call centers and helpdesks being the most
common. Many nations can offer language-based services ÔÇ“ provided that reliable, fast,
and cheap telecommunications is available. Many thousands of call-center jobs have been
transferred to English-speaking countries, such as India or the Philippines. Mexican call
centers are serving the Spanish-speaking customers of American companies. China is
offering Japanese-speaking staff to service Japanese ´¬ürms. Other nations have been slower
to grow their exporting of IT-enabled services. South AfricaÔÇ™s call-center industry suffered
from underexposure in the international marketplace. The nation has some clear advan-
tages for European users, such as cultural similarities, common time zones, and political
stability. The industry awoke to the opportunities as evidenced by the aggressive market-
ing overseas by the Western Cape Province, home to Cape Town. Governments can help
grow customer interaction services by facilitating training in languages, accent neutraliza-
tion, accent comprehension, telephone etiquette, listening skills, and problem/objection-
Back-of´¬üce operation services
Some of the back-of´¬üce work has very low-entry barriers in terms of skills and can be
done in many developing countries. One of these services is the handling of reserva-
tions. This is a high-volume, low-value type of work where competition is based on
206 Other stakeholders
factory-like approaches and low wages. Thus, American hotel chains have turned to
Jamaica and British Airways, in 1996, turned to India.
Other services require some speci´¬üc knowledge. Claims processing for medical insur-
ance companies allow doctors and hospitals to pass on all relevant records (via scanned
images) to the offshore staff which handles all the paperwork. This work requires knowl-
edge about the details of medical insurance. Understanding medical words and phrases
is also required for medical transcriptions. The conversion of dictated voice recordings of
doctors to paper is a niche market for specialized offshore service providers.
Document imaging also allows offshore locations to perform ´¬ünance and accounting
services, such as processing accounts payable, accounts receivable, ´¬ünancial reporting,
tax consulting, or internal audit services. Speci´¬üc domain knowledge is required for
these functions, and some countries possess these skills. The Philippines has a large sup-
ply of accountants trained in US accounting standards. US-based Procter & Gamble per-
forms its accounting services in the Philippines, where 650 employees complete the
corporationÔÇ™s tax returns for its global operations.
Data and content integration
Data entry is required when data, stored on paper, needs to be converted into digital
form, and scanning is not an option. In those circumstances, manual data entry is a
solution. Data entry requires the lowest level of computer literacy and does not require
strong language or functional skills. It also requires very little interaction between the
customer and the offshore provider. It is therefore a service many Tier-2 and -3 coun-
tries can offer. An example of data entry is the of´¬ücial Dutch dictionary, which consists
of 40 volumes. When it required migration to digital format, the books were sent to
India, and keyed in twice by different groups of typists. The two data sets were auto-
matically compared to ´¬ünd any differences and then corrected.
Digitizing is the process of converting texts, images, video, or records into digitized
forms. The advantages are long-term preservation of the documents and easy access to
the information. It is a labor-intensive activity, and ideally suited to be done in a low-cost
country. The imaging is sometimes done on the clientsÔÇ™ premises. In other cases, the orig-
inal records, or copies, are shipped offshore. Anglo-Dutch publisher Reed Elsevier is
digitizing all its scienti´¬üc publications, printed over centuries, at SPI in the Philippines.
A Swiss newspaper is digitizing its historical publications at Dakor, in North Korea.
Geographic information systems (GIS) are increasingly used by governments,
telecommunications, and cable television companies. Much of the worldÔÇ™s geographic
data still resides on paper sources and needs to be converted. A large proportion of the
cost of GIS is in data collection, interpretation, and conversion, making offshoring a
viable option. GIS providers can also be found in several Tier-2 and -3 countries. For
example, one of the leaders in this niche in India is Rolta with more than 2000 employ-
ees. Rolta is also moving to higher value activities by expanding into technical serv-
ices, such as CAD modeling, plant design, and mechanical design.
207 Building software industries in developing nations
Another area of IT-enabled services growth is that of animation services. Animation
(e.g. cartoons) is very labor intensive to produce, even when computers are used, and
is therefore ideal for offshoring to low-cost nations. These services include the pro-
duction of a scenario, a storyboard, the drawing of character models and backgrounds;
as well as the intermediate process of layout creation, animation celluloid, scanning,
coloring, and composition. Countries such as India and the Philippines are offering
animation services. Interestingly, North Korea, with very low labor costs, has targeted
animation. SEK Studio in Pyongyang is one of the largest animation studios in the
world with an artistic staff of 1600. It produces for French, Italian and Spanish ´¬ülm and
Why developing nations should invest in building a
software export industry
Software exports demonstrate the advantages of globalization without most of its nega-
tives. The industry is a clean, non-polluting industry employing people using their
brains rather than their muscles. Software is also a general-purpose technology that is
used across all industries and therefore has positive impacts across the economy. Policy-
makers in developing nations recognize that such high-tech exports are more valuable
to their economies than exports of coffee, or minerals, or assembled goods, or tourism.
In this section we introduce some of the advantages of a software export industry,
namely the creation of jobs, revenue generation, improvements within organizations, and
other positive impacts on society.8 The latter bene´¬üts are spillover effects, or as economists
refer to them, externalities. Spillovers are the bene´¬üts outside the software-exporting sec-
tor itself ÔÇ“ to other businesses and to other aspects of society. In summary, the software
sector (domestic and export oriented) is important to national well-being because it has
multiplier effects: it is a general-purpose technology and because it creates spillovers.
Many of the examples in this section are from India, where the impacts of the soft-
ware export industry have been relatively most dramatic.
Unemployment is a major and persistent problem for all developing countries with
insuf´¬ücient opportunities for educated youth. Sadly, acquiring a university degree is no
guarantee of a job. Due to unemployment, many countries experience labor migration,
resulting in a brain drain. To address this dire situation, the promise of the software
export sector is job creation for the nationÔÇ™s educated citizens. Nations have been look-
ing with envy at India, where the software export sector offers direct employment to
more than 260,000 people in 2004 and ancillary jobs are at least double that number.
In the Tier-2 and -3 countries, the amount of software jobs is still modest, but growing.
For example, the Vietnamese IT sector, which we described in Chapter 4, employed
208 Other stakeholders