The government is rolling out a national information infrastructure from the top down. It's doing this because it believes this will prove an extraordinarily beneficial tool in the administration of China, and if there is one thing the communist party is serious about, it is running the country.

At its heart, China's leadership quite clearly believes that the growing convergence of computers and telecommunications, and the potential this offers to create a networked society, will benefit it by allowing it to strengthen its ability to exercise administrative control over the entire country. Much of the evidence points to it being right in this belief. While information technology has allowed a certain amount of decentralisation in where work is performed, it has done so by abolishing the distinction of distance. It does not matter where, for example, you are in the world, you can still trade shares in New York; however, in order to do so, you have to play by the rules, or protocols, of the authorities which allow you to participate.

Doing this clearly suggests that the information networks offer the potential for power structures to be reinforced rather than weakened. After all, governments that have moved to embrace the emergence of the Internet remain the repository for its administration. The one counter argument of strength is that an increased concentration of power at the centre requires a greater ability to process information and, in general, the human resources are not available to handle and interpret the data collected. However, it does not necessarily follow that the difficulties in managing the people required to process information contradicts the strengthening of administrative hierarchies.

One view has it that communication networks have a vast but contradictory potential. For example, new information technologies can make both markets and hierarchies more efficient. The increased efficiency is mostly born out of the considerable reduction in coordination cost that the new information networks can provide. Thus it could be possible that a reduction in coordination costs could increase the proportion of economic activity coordinated by markets, which are more communication-intensive than hierarchies. But market growth leads to an increase of economic actors which in turn requires an increase in the amount of communication required by the market to operate efficiently. To the leaders in Beijing this sounds like an increasing demand for administrative oversight, not a lessening of their role.

Even research suggesting that information technology and networks have weakened hierarchies and central control point towards an acceptance of shared standards. For example, what were traditional, centralised market places, say for financial products, are being replaced by electronic market places accessible from anywhere. Certainly then, the transition to an information-based economy can result in the shift from traditional functional hierarchies to informal networks.

To emphasise the informality and geographic dispersal allowed by networks however is to miss the point about their minimum requirement in order to function. To operate, any network must have agreed standards and protocols for communication and business to take place. If standards are controlled or enforced from a centre, they at the very least give the centre a greater say over the form in which any specified activity takes place. To return to our example of share trading on the New York Stock Exchange: just as the increasing access allowed by new information technologies has increased economic turbulence and challenged the role of financial intermediaries (such as the brokerages), so too it has increased the importance of the US Federal Reserve in creating stability and leadership.

The Chinese government's motivation for building out a national information infrastructure is twofold: to promote economic development and to exercise greater administrative control across the country. From the very start, it has believed that extensive wiring of China combined with widespread incorporation of information technology could increase central economic coordination and administrative control while at the same time achieving improved productive efficiency and performance.

This has a lot of appeal, as one of the most profound outcomes of Deng Xiaoping's economic reforms has been an unwanted but sustained decentralisation of political power. With this has come an increasing tension between the centre and periphery, manifested most clearly in Beijing's ongoing struggle to rein in Guangdong's freewheeling economic ways. The economic growth in the province over the last two decades has been stunning. Less welcome by far has been the discovery of large amounts of hitherto unrecorded foreign debts at Guangdongbased companies, such as the now bankrupt Guangdong International Trust and Investment Corporation. But this has merely served to underline to the centre the importance of having adequate machinery in place to record and track business developments in distant parts of the country. New information technology and communication networks have seemed to offer the tools to achieve both the expansion of market forces which should allow the economy to grow most optimally, while allowing central control and monitoring of how this is being achieved.

China's first move in the direction of building out a national information infrastructure was the initiation of what were known as the Golden Projects. These were, and are, an ambitious information infrastructure initiative aimed at simultaneously developing an information economy and building administrative capabilities. Their administrative relevance was reinforced in December 1993 through the creation of a high-level leading committee under the State Council, known as the Joint Committee of National Economic Informatisation. When officially launched the following year, the projects had three overarching goals:

These three goals overlap but are significantly different in nature. The first is concrete -- an information infrastructure over which data can flow. The second is emotional -- a cry for the country to become a modern nation. The third is a combination of goals one and two: it is clearly the most important, and the driver of the other two. Realising it will allow the central government to reacquire administrative control by being able to act as an information gatekeeper for the country.

Initially, the Golden Projects comprised three elements: Golden Bridge, Golden Card and Golden Gate.

Since the initial three golden projects were proposed, a series of other programmes have emerged and been implemented. While the agenda of the initial projects was the roll-out of information networks, these later projects have generally involved user-driven applications designed to use the information infrastructure to promote economic reform. With most of these Golden applications, government agencies are attempting to both increase economic efficiency and centralise control of information.

Two are particularly worthy of note. The Golden Sea Project, despite receiving little publicity, exemplifies the leadership's interest in information networks. It interconnects China's top government leaders and provides them with immediate access to reference data from other institutions, organisations and offices under the direct jurisdiction of the Communist Party Central Committee. Launched in 1994, it was the first Golden Project to be finalised (it was up and running by 1995). Among the organisations it brings together are the State Statistical Bureau, People's Bank of China and State Information Centre.

The other is the Golden Intelligence Project, better known as CERNET: the China Education and Research NETwork. CERNET was the vehicle for the Internet's entrance into China. Initially funded with Rmb80 million (US$9.6 million) of seed money from the central government, it will ultimately interconnect all campus networks with each other and into the international Internet. There is a certain irony in the government allowing the very group of people which fomented the pro-democracy movement of April to June 1989 -- the academic community in general, and students in particular -- to be the first to gain access to the Internet.

In addition to CERNET, three other data networks in China offer direct Internet access: China Science and Technology Network (CSTNet), Golden Bridge Network (GBNet) and China Telecom's ChinaNet. CSTNet is similar to CERNET, but smaller in scale: it connects subsidiaries of the Chinese Academy of Sciences. GBNet is the "other" network, run by JiTong, a state-owned company formerly linked with the now abolished Ministry of Electronics Industry.

Which leaves ChinaNet, by far the most important of these four, run as it is by the operator of China's national public telephone network. Often referred to as the 163 network after the number users dial to gain access to it, it is the only part of the Internet in China with a legal Internet protocol address and a formal Internet domain name. This means it controls who can set up web sites and which web sites can be blocked. Although technical means do, and will always exist for users with the know-how to find their way to blocked sites, fundamentally this means that China Telecom remains in overall charge of who accesses the Internet via the public telephone system.

Only ChinaNet and GBNet can sell Internet on commercial terms to other Internet service providers (ISPs), the companies which in most other parts of the world act as the link between a user and the Internet. Because China Telecom has the vast majority of the infrastructure in place, this in effect means it is the monopoly supplier. All the 150 or so of China's ISPs are small and local, and China Telecom has shown no compunction to date in squeezing as much money from these businesses as possible, with the result that whereas in the United States, line rental only account for 5% of an ISP's costs, in China the average is nearly 80%.

Even more restrictive is a China Telecom practice of linking line rental to the amount of revenue per line. Consequently, instead of rental declining with volume, it rises, making an ISP less profitable the more it increases its user base or usage. Given a playing field tilted so steeply against them, most independent ISPs have not only found it impossible to compete with ChinaNet, they have found it impossible to stay in business without receiving some degree of assistance or lenience from China Telecom. As a result, although China saw a small blossoming of ISPs in 1997 and 1998 (many being small bulletin board service operations which decided to go commercial), many companies granted ISP licences have stopped offering ISP services, or have been incorporated into the ChinaNet framework.

As a consequence, the majority of subscribers, whether companies, organisations or individuals, are connected with ChinaNet, either directly or indirectly. Educational users will of course access the Internet via CERNET, and those in the Chinese Academy of Sciences via CSTNet. The Golden Bridge Network, although offering Internet access to individuals, is still essentially a network in development and is being developed principally to service corporate accounts.

So, much as many people might like to think the Internet is part of a bottom-up explosion of individualism in China, it is not. It is instead a highly centralised network, largely running through a single carrier, China Telecom, and a monolithic infrastructure, China Telecom's public network, to carry data services to all other Chinese and foreign ISPs, companies and individual users.