The Internet in India and China

It is our hypothesis that the Internet will make a significant contribution to the quality of life in developing nations. Since roughly 38% of the global population lives in China or India, the progress of the Net in those nations deserves careful study and consideration. We conducted a study of the state of the Internet in China and India during the summer of 1998,[1] and found much that is typical of developing nations in both.

The paper begins with a comparison of the state of the Chinese and Indian Internets along six dimensions. India joined the Internet before China, but the Chinese Internet quickly caught or surpassed India on each of our dimensions. We offer explanations of this situation by comparing determining factors in telecommunication infrastructure, human resources, equipment and the economy, and government interest and support.

In spite of frequent claims that the Internet will erode national sovereignty, government interest and support is seen to be important both directly and indirectly through its influence on the other factors. Both governments are now committed to the Internet, and we examine changing government roles, primarily the adoption of an ambitious Action Plan in India and the consolidation of two key Ministries in China. The Indian Action Plan addresses each of our six dimensions, and is designed to elevate India to the level of IT superpower. The impact of Chinese consolidation is less clear, but will also be important in determining the future of the Internet there. We conclude with a discussion of sources of uncertainty.

Comparing the Internet in India and China

We evaluated the Indian and Chinese Internets on six dimensions, starting with pervasiveness. The Internet is more pervasive in China where there are an estimated 1.2 million accounts versus only approximately 200,000 in India, and host count estimates give China a lead of roughly eight to one.[2] The research and education networks have been particularly effective in China, where they account for more than half a million users.

China also leads in geographic dispersion. Commercial Internet access is available in over 200 cities representing all Chinese provinces, while India has Internet POPs in only 17 of 32 States and Union Territories. China qualifies for a higher rating on our scale, but, usage is concentrated in large cities in both nations, and villages (roughly 70% of both populations) are completely unserved. India has opened the ISP market, which was a government monopoly at the time of our study, and may therefore improve rapidly.

China leads in sectoral absorption as well. Business connectivity is rare (under 10%) in China, but there are fewer than 400 connected businesses in India. While connectivity is almost non-existent in primary and secondary schools in both nations, over 300 Chinese universities and 200 research institutes have direct connectivity. Government connectivity and Web sites are rare in both nations as is usage in the health sector.

Connectivity infrastructure is a function of domestic backbone, the prevalence of high-speed access, Internet exchanges and international bandwidth. India has little terrestrial backbone, relying almost exclusively on satellite links. China uses both satellite and terrestrial links. For example, ChinaNET connects its centers with 155 Mbps circuits and connects to its 200 POPs at between 2 and 34Mbps. Nothing close to this exists in India. Neither nation operates Internet exchange points at present, but China has plans to.[3] China has more organizations connecting with leased lines, and is experimenting with cable modem and xDSL, but they are not deployed in either nation. Finally, China has more than double India's international bandwidth. In spite of China's relative advantage, we must bear in mind that aggregate bandwidth per user is very low compared to a developed nation, rendering interactive applications such as Web access impractical in many cases -- email is the primary application in all developing nations.

Organizational infrastructure is concerned with competition in the telecommunication and Internet industries and with coordination and organization in the Internet industry. Telecommunication is monopolized in both nations. Ironically, there has been more Internet competition in China where there are four interconnecting networks. Two of these serve only education and research, but the other two are open, and there are 200 competing access networks downstream from these. Until recently, the Indian government monopolized the Internet, but backbone and access competition have now been authorized. While interconnecting network competition is beginning in India, the ministries which operate China's commercial interconnecting networks, the Ministry of Electronic Industries (MEI) and Ministry of Post and Telecommunications (MPT), are being merged into a new Ministry of Information Industries (MII), which may reduce backbone competition. At the same time, there is immense pressure on the part of the central government and Premier Rongji to introduce competition into the telecommunication industry. As part of this attempt, China UNICOM, the nations second telecommunications provider, is being reinvigorated. During the latter part of 1998, a number of plans for restructuring China Telecom have been considered by the Chinese State Council. The one most likely to succeed involves separating the paging, wire-line, and mobile communication businesses, and breaking the wire-line business into 18 separate companies along regional lines. What impact this restructuring will have on the competitiveness of the Internet backbone market remains to be seen.

Sophistication of use is comparable in the two nations, with the Internet increasing efficiency of conventional organizations and processes, for example in substituting for mail and FAX. Both nations may make similar innovations in the future since they are demographically similar in many ways. For example, both have large rural populations, and will be motivated to innovate in the use of the Internet to address the needs of villages and in inventing new applications, technology and organizations to enable that service. (The Indian government has officially recognized this as a priority).

In summary, Table 1 shows that, today, China is currently doing as well or better than India on each of our dimensions and key sub-components. This situation may change with the advent of Internet competition and the government adoption of an ambitious IT Action Plan in India. We will discuss this changing government role below, but first, why did China start late and how did it surpass India?

Dimension or component	Advantage: C=China E=Even
Pervasiveness
users	C
hosts	C
geographic dispersion
top-tier political divisions with POPs	C
number of cities with POPs	C
sectoral absorption
commercial	E
education	C
government	E
health	E
Connectivity infrastructure
domestic backbone	C
high-speed access	E
exchanges	E
international bandwidth	C
Organizational infrastructure
telecommunication competition	E
backbone competition	C
access provider competition	C
coordinating organizations	E
Sophistication of use	E
Table 1: Dimension Comparison Summary

Comparing Key Determinants

Internet progress in a nation depends upon determining factors including the availability of telecommunication infrastructure, networking technicians and trained, demanding users, networking and end-user hardware, and an interested, supportive government.[4] China enjoys advantages in many of these areas.

Telecommunication Infrastructure

Telecommunication infrastructure is an important element in Internet growth, and, although India and China lag behind industrial nations, China has invested heavily during the 1990s. As a result, Chinese telecommunication infrastructure has grown much faster than that of India. Table 2 shows this growth gap for selected indicators, and virtually all aspects of China's telecommunication infrastructure have grown more rapidly. The gap has continued to widen since 1996, as India's capital expenditure for telecommunication was $1.794 billion in 1996 while China invested $13.038 billion.[5] This rapid buildup in China requires outside financing, and China has been much more successful than India in attracting direct foreign investment.[6] The Chinese may not have been thinking of the Internet when they decided to invest in telecommunication infrastructure, but it has helped regardless.

	China		India
	1990	1996	1990	1996
Mainlines (1,000)	6,850.3	54,947	5,074.7	14,542.7
Mainlines/100 capita	0.6	4.46	0.6	1.54
Mainline waiting list (1,000)	688.8	812	1,961	2,894.2
International circuits (1,000)		30.8		11.5
Outgoing international (M minutes)	350	1433.2	146.7	342
Table 2. Indicators of telecommunication growth and capacity[7]

The Chinese also operate their telecommunication system more efficiently. They have over 4.4 times the revenue per employee and 3.3 times as many mainlines per staff member as India.[8] One explanation of Chinese efficiency is that they decentralize decision making, pushing investment and service decisions to the provincial or city level.

India's legendary bureaucracy also cuts efficiency. For example, spectrum allocation is parceled out to government agencies each of which manages its own slice, leading to sub-optimization.

Human Resources -- Technicians and Users

A robust Internet also requires investment in human capital. Both networking leaders and technicians and demanding users are required. India and China followed a common pattern in that the Internet began with universities and research institutes.[9] In such cases, the early networking community provides leadership and technical knowledge.

While Chinese academic networking began with X.25 in 1987, IP connectivity did not begin until 1993, and broader university Internet access (both UUCP and IP) began to take off in 1995. The two academic networks, CERNET and CSTNet, were instrumental in extending the Internet beyond the academic and research communities by their example and leadership and in offering commercial service, and they have continued to thrive, accounting for roughly 50% of Chinese users. Approximately 25% of Indian users are from universities.

The education and research network (ERNET) pioneered Indian networking, but did not play as important a technical and leadership role. They were funded by United Nations Development Programme (UNDP) for many years, and were therefore isolated from performance and market demands. They spent early years planning at a time when UUCP networks were being rapidly deployed in other developing nations, and when available, their service was often unreliable. When UNDP funding was withdrawn in 1995, it was not clear that ERNET would even survive.[10] They had interim funding for two years and finally became an Autonomous Society in 1998, but their budget, staffing levels, telecommunication-subsidy levels, and office location are yet to be determined.

Chinese universities were more entrepreneurial than those of India. Funding in China came from both the Ministry of Education and the individual universities, while Indian universities received very little money from the government and paid nothing themselves.

Network leaders and technicians are on the supply side, but the Internet also needs trained, demanding users. These are initially from the university and research communities, but demand from other sectors soon becomes dominant. At this early stage, a lack of demand does not appear to have constrained Internet diffusion in either nation. While education and literacy rates are low (particularly in India), the absolute numbers of potential users are high. Roughly five percent of the population of South Asia speaks English.[11] This translates into nearly 50 million potential Indian users since English speakers are generally well educated and able to afford a computer and Internet account or access at a public venue. The educated middle class in China is also large in absolute terms. Both nations have large university systems -- in 1990, 4,425,247 Indians and 2,651,396 Chinese were enrolled in higher education.[12] For the time being, the number of educated, middle-class people in both nations is large relative to the number of Internet users.

Equipment and the Economy

The Internet also requires end-user equipment (PCs), and networking equipment. Relatively large numbers of PCs are a direct advantage in that they are available for use and an indirect advantage in that the people operating them become trained potential users. The ITU estimated that there were 3.7 million PCs installed in China and 1.1 million in India in 1996, [13] and, as shown in Table 3, the gap is expected to grow.

	China		India
	Millions of PCs Sold	Annual Growth Rate	Millions of PCs Sold	Annual Growth Rate
1996	2.98		0.38
1997	3.9	31%	0.47	24%
1998 (estimate)	4.95	27%	0.62	32%
1999 (estimate)	6.27	27%	0.85	37%
Total	18.1		2.32
Table 3. PC sales (1998, 9 are estimates)[14]

The Chinese advantage in PCs reflects two factors. China has been more heavily involved in PC manufacturing and trade and has established brands like Great Wall and Legend. China produced over 1.5 million PCs in 1996 and the government predicts over 10 million units by 2000. One company, Legend, produced 350,000 PCs in 1997, while all of India assembled only 500,000. China also produces and exports PC components, for example, 12 million motherboards and 8 million monitors in 1996.[15]

The second factor is affordability. Whether a computer is purchased by an organization for employees or by a family for the home, the cost is large relative to budgets in developing nations and hard currency is required. A low-end, Internet-ready PC costs around $1,000 in either nation, but China is richer, making PCs relatively more affordable.[16]

Networking equipment is also an issue. To date, it has nearly all been imported in both nations; however, China seems more likely to eventually develop a network equipment industry. One clue to the possible future differences between the two is provided by the fact that the Chinese telecommunication-export industry was more than 25 times as large as that of India in 1996. [17] Cisco Systems has agreed to assemble routers in China. This may have been partially motivated by pressure from the Chinese government, but access to infrastructure, local availability of components like power supplies and service like sheet metal fabrication make China an attractive manufacturing location. Cisco has decided to invest $100 million in various activities within China.[18] Bay Networks (now part of Nortel) will establish an architectural lab in Beijing,[19] and 3Com will invest $100 million in several activities, including networking.[20]

It is interesting to note that Cisco has opened a 75-person software development center in India, but this reflects a smaller investment.[21] It parallels the effort of the Microsoft Windows NT development group in Hyderabad with roughly 20 programmers.[22] The sizes of these programming groups and the hardware manufacturing volumes in China, indicate that the hardware side of the “information economy” may have a greater impact on GNP than the software side.

All of the above factors -- telecommunication infrastructure, human resources and equipment require costly investment which must occur within the context of the broader economy. While both India and China are developing nations with scarce resources, China is more productive and prosperous today (Table 4). In 1980, GDP per capita (in 1987 dollars) was $134 in China and $262 in India.[23] Indian policy changes in 1980 helped GDP growth accelerate to over 5% per year during the 1980s, but since China's market-oriented reforms began in 1978, real GDP growth has averaged over 9% per year.[24]

	India	China
GNP ($ billion), 1995	325	745
GNP per capita, 1995 ($)	340	620
Gross domestic investment as % of GDP, 1994	23	42
Gross domestic savings as % of GDP, 1994	21	44
Imports as % of GDP, 1994	12	24
Exports as % of GDP, 1994	15	23
Direct foreign investment, 1995 ($ billion)	3.487	35.849
Table 4: National income and trade indicators.[25]

In addition to being more productive, China has greater propensities to save, invest, and trade than India. Since the time of Mahatma Ghandi, India has followed a self-reliance policy motivated by economic and security considerations. Reform and opening began in 1991, and direct foreign investment rose from $162 million in 1990 to $1.3 billion in 1995,[26] but China still enjoys a wide, growing lead in international investment and trade.

Government Interest and Support

While telecommunication, including the Internet, may erode some aspects of national sovereignty, all of the factors we have mentioned are influenced directly or indirectly by government policy and action. In every nation, government plays a central role in Internet diffusion, but that role can change over time. For example, the U. S. government led in establishing the Internet and earlier networks through purchases, research, and direct operation, but it has drawn back as industry stepped in.[27] Historically, the Indian government has impeded the Internet with a combination of indifference and government monopoly, though it appears ready for a bold change of direction, and the Chinese government first considered the potential risks, then moved relatively rapidly once it decided to join the Internet.

Indian IT policy has vacillated over the years. The deployment of mainframe computers was interrupted when IBM withdrew from India in response to a 1976 law limiting foreign ownership of business to 40 percent. Rajiv Gandhi assumed leadership after the assassination of his mother in 1984, and identified telecommunications and information technology as a "core sector" along with traditional industries like power, steel, oil and automobiles.[28] The Internet languished after Gandhi's assassination with the government using the authority granted in the Indian Telegraph Act of 1885 to stop private ISPs from operating. Only powerful government agencies have been able to become ISPs serving limited constituencies, and the Ministry of Communication has kept a monopoly over commercial ISP service. Whether motivated by a desire to maintain revenue, power, or control over technical issues, the Indian Government moved slowly in embracing the Internet, and resisted the global trend toward privatizing telecommunication and introducing competition.

The election of the BJP in 1997 signaled renewed interest in IT and the Internet. BJP advocates economic liberalization, and listed IT as one of the government’s five top priorities. This commitment led to the establishment of a high-level IT Task Force in May, 1998. If the recommendations of the Task Force are implemented, the Indian Internet will grow dramatically, and may close the gap with China.

Governments are aware of the potential advantages and risks of the Internet.[29] It is seen as having potential to improve the economy, education, health care, environment, and the quality of life, particularly in rural areas, but these are weighed against the possible threat to the stability of the regime,[30] national security, and cultural values.

While the Indian government appears to have been slow to act out of bureaucratic indifference, other threats do not seem to have played a major role. This relative lack of concern is consistent with India’s long-standing commitment to a free and often highly critical press, and Indian pride in being the world’s largest democracy.

China is more concerned than is India about the political risks of information in general, including the Internet. The Chinese government is sensitive to the impact that the mass media, facsimile machines, tape recorders, and Internet news groups had during the Tiananman Square demonstrations. It also witnessed the fall of the Soviet Union after information flows were liberalized, including the use of the Internet during the Soviet coup attempt.[31] Ironically, Chinese national security concerns led the government to focus its attention on the Internet prior to India, and this may have accelerated growth.

While it took some time to decide to what extent they would participate in the Internet and which agencies should be involved, the Chinese decided the benefits outweighed the risks. Once they decided to move, they acted quickly. The Chinese State Council authorized four government organizations to run networks that interconnect with the global Internet, and the threat of competition between them influenced deployment, service, and pricing. In addition to running their own ISPs, both China Telcom's ChinaNET and the Ministry of Electronic Industries (MEI) ChinaGBN were willing to sell support to private ISPs; whereas, in India there was no backbone competition.

While the Chinese government has moved forward with the Internet, it is attempting to minimize the risks by maintaining control over access, use, and content. The December, 1997 regulations promulgated by the Chinese Ministry of Public Security are extremely broad and forbid the transmission of information that injures the reputation of state organs, incites division of the country or resistance to the Constitution, laws, or implementation of administrative regulations. There are also prohibitions on promoting feudal superstitions, sexually suggestive material, gambling, or violence, and user registration is also mandated. The recent sentencing of Lin Hai to two years in prison for providing 30,000 email addresses to a “hostile foreign organization” and the establishment by the national police ministry of task forces for Internet monitoring in search of “seditious talk” are indications of China’s determination to control the Internet.

To some extent, the differing levels of concern with security issues between China and India may be a result of the influence and size of their respective militaries. Chinese Military expenditures are nearly four times those of India,[32] and the PLA has been more heavily involved with the Internet (seeking to be an ISP) than the Indian military.

In contrast to the Union Government, Indian State governments have had a positive influence on the Internet.[33] For example, in West Bengal and the capital of Andhra Pradesh, Hyderabad, proactive State governments have provided incentives to involve private firms in Internet-related ventures, in which both the government and the firm share equity, establishing a Singapore-like middle ground between government control and laissez faire. In these and other cases, state projects and partnerships between state and local governments and private enterprise have preceded and contributed to national planning and action in India.

Similarly, decentralization has played a role in China. The distributed nature of the Internet has worked well within the context of the distributed organization of China Telecom itself. Posts and Telecommunications Bureaus at the city level can invest in their own ISPs while being connected to the provincial or national backbone. This led to a rapid rollout of ChinaNET during 1995-6 to all Provinces. China has embraced markets, and there is considerable competition among local governments in IT and other areas.

Looking to the Future

The governments have played major roles in both nations. The Chinese government addressed the Internet late, after their own academic networks and world developments made it clear that this was strategic infrastructure. Once they decided the potential benefits outweighed the risks, they created two backbone providers and many access providers, while attempting to maintain control over access, use and content. The Indian government has been more passive, neglecting the Internet and allowing it to be controlled by bureaucratic monopolies.

However, the situation in both nations is now changing dramatically. Coincident with the merger of the MEI and MPT into the MII, the Chinese government separated the regulatory authority from the operating organizations, and is seeking to introduce some degree of competition. It remains to be seen what the impact of this move will be, but it will surely affect the course of Internet diffusion in China.

The changes in India are potentially more dramatic. The speed with which the IT Task Force moved is indicative of the changing government attitude. Within 90 days, it produced an extensive Background Report and a 108-recommendation Action Plan.[34] The Task Force could act quickly because it built upon the experience and frustration of the state governments, universities, and industry. Much of its plan is also consistent with the thinking and recommendations of international bodies like the WTO, ITU, and World Bank, and it had the example of similar plans in Singapore and other nations. The Task Force did not start from scratch -- it surveyed the opinions of Indian computing and networking leaders and recommended what it heard. This was less a task of invention than of sparking action on a consensus that had already evolved.

The remainder of this section compares the likely impact of the Indian liberalization and the Task Force recommendations and Chinese consolidation and other factors on the diffusion of the Internet in the two nations. We consider the six dimensions introduced earlier.

Pervasiveness

At present, the Internet is confined to large cities in both nations. While these groups are far from saturated, high levels of pervasiveness will require service to the lower urban classes and villages, which raises issues of public access, service in villages, education and language, and affordability.

It will be many years before PCs are affordable by average Indian or Chinese families, and the bulk of urban residents will be dependent upon public access facilities. The Indian Action Plan calls for the establishment and facilitation of public access points, and it suggests that Internet access might be provided by upgrading the 600,000 privately run public telephone offices in the country today. These will not solve the entire problem, because many are small and under-capitalized and staffed, but some will doubtless become Internet access points. Some state governments also have plans for public access programs. While the Chinese have no policy on universal access or the spread of the Internet to villages, they do plan to extend telephone service to every village, indicating that universal access is a consideration.

Pervasive penetration will require connectivity in villages, which comprise approximately 70% of the population in both India and China.[35] The Indian Action Plan calls for taking "all the necessary steps to boost IT for agricultural and integrated rural development," and calls for pilot and research projects. These are lofty goals, but telephone service in rural India is very poor, with approximately half of villages having no phones. The Indian government may be reversing a policy of neglect, and the government appears to be more aware of villages than that of China.

Whether in villages or urban areas, computer users must be literate, and illiteracy will constrain Internet penetration in both nations.[36] The problem is greatest in India, with an overall illiteracy rate of 48% as compared to 19% in China.[37] This gap may be expected to continue into the future, since the average primary school class in China has 22 students per teacher compared to 64 in India.[38]

Language is also a major issue in achieving widespread network utilization. The Indian constitution recognizes 14 languages, but an estimated 179 languages and 544 dialects are used. While 30% of the population speaks Hindi, it is still an unpopular, foreign language in the south, Bengal, and other regions.[39] China is dominated by the Mandarin dialect, and both Cantonese and Mandarin use the same character set. This has made standardization on a native language in China easier than India, leading to a higher level of Chinese content on the Internet.

The fact that most e-mail packages did not support 8 bit encoding, made using English a necessity at first, but today, as programs that support Chinese become common, more and more Chinese are utilizing Chinese in email. There has been a government effort to develop Chinese-language content.

In India, the local language market is fragmented, and English is therefore used on most Web sites; however, the Action Plan states that "A major promotional campaign shall soon be launched to boost IT in Indian languages." In addition to developing content, technical issues of support for Indian languages in software packages and font standards must be addressed. The government has supported some research in this area[40] and Windows 2000 will support Hindi and Tamil. ERNET also distributes local-language software free of charge.

While government action and programs will encourage Internet penetration in both nations, free market purchase of PCs will still be a key determinant of pervasiveness, and this is tied to affordability. The Indian Action Plan will help with duty reduction and other incentives to lowering computer prices, eliminating gift taxes on PCs, financial subsidy for teachers and students, etc.; however, both nations will benefit from the rapid reduction in the cost of Internet-capable PCs arising fr