From the Swiss Review of Political Science, Spring 1999:

Internet Governance and the Ancien Regime

By Kenneth Neil Cukier

Senior Editor, Communications Week International

Paris, France

* * *

"We are creating the most significant new jurisdiction we've known since the Louisiana purchase, yet we are building it just outside the constitution's review".

-- Larry Lessig, June 10, 1998.

"Ultimately, Larry wants to make cyberspace safe for law. I want to keep law out of cyberspace".

-- John Perry Barlow, August 24, 1998


When the Chinese government named Tung Chee-hwa to replace Hong Kong's UK-appointed governor Chris Patten in June 1997, they chose a very post-modern title for the political leader: chief executive. It might have resonance with Western investors, and certainly it reassured them that despite the political changes, the raison d’etre of Hong Kong as the commercial entrepot for Asia remained constant.

And it falls in line with a broader trend in governance and international relations. Just as governments around the world have replaced diplomacy with trade policy -- representing a historic shift in the notion of national security from the military to the economic realm -- so too have they substituted traditional government activities with market-based approaches. If the UK's National Health Service can outsource all citizens' medical records to a private company, Electronic Data Systems Corp. of Plano, Texas, perhaps it makes sense for government itself to outsource its functions to private entities. We, the children of Adam Smith and fans of Francis Fukuyama, are all free-market democrats now.

Or are we?

Larry Lessig does not rue the developments in technology and society that has led the U.S. government to bow out of running the domain name system and other key aspects of the Internet's infrastructure. He's too smart to be distracted by sentimentalism; and besides, he supports the move albeit with pragmatic and cogent qualifications. Rather, Professor Lessig's analysis is deeper, and his objective a philosophical one. Without prejudice, he questions why Americans have become so supp'd full of horrors of government that they deny its applicability to serve the public interest, and specifically in cyberspace.

He's got a good point. Non-profit entities in positions of public trust to uphold the popular will -- which is the Internet Corporation for Assigned Names and Numbers (ICANN) -- basically describes the role of democratic government. So why has "private-sector leadership" or "industry self-regulation" been accorded a status as sacrosanct as motherhood? Cyberspace has real consequences in the space we live in, Lessig rightly explains. Shouldn't government, the guarantor of constitutional law, have its place in there too?

Yes and no. Professor Lessig's perspectives on "governance" unfortunately amalgamate two very distinct things: the "governance of the Internet" and "Internet governance." This is not a silly game of semantics; the two notions are wholly different. It might be restated as: How the Internet is used, as opposed to how the Internet self-organizes to technically work. Lessig makes no separation between regulating the things that happen on the Internet compared with how the Internet itself is structured. It is as if one were to lump together the telephone network's operational signaling system with the conversations themselves simply because both travel along the same copper wire during a call.

However, this distinction must be made to fully understand the issue of government and the Internet -- what its potential role might be, where it is best suited to act, and where it isn't. To do otherwise, risks handing the control of the new medium over to the ancien regime, governments, which could have disastrous consequences for the Internet's future development.


The first area, the "governance of the Internet", entails how traditional matters of public policy are naturally extended to cover the Internet. For example, government has long regulated speech by placing limits on material that is harmful to minors, and it is unsurprising that many nations have passed laws to apply these standards to Internet content. In other areas, like intellectual property rights and consumer protection, there is a deep history of government regulation that is now being transferred to cyberspace. Governments have long raised revenues through taxation -- and intend to do so over the Net as well. And new challenges have appeared, such as privacy, that force nations to "govern the Internet".

In this case, "govern" means to bring order to the Net, not because the Internet is chaotic per se, but because the international nature of the medium creates a vacuum for national regulations that governments feel justified in filling.

Why shouldn't they? Governments already regulate these areas in the off-line world, so cross-applying these regulations to the Internet domain is only natural. Importantly, all of these issues represent the "superficial" aspects of the Net, insofar as they concern how the Net is used and the content that rides atop the infrastructure.

Regarding these issues, Lessig's views are tremendously insightful. He is the first to appreciate how the phenomenon of industry self-regulation is representative of a broader mistrust of government so profound that people are prepared to genuflect before technological solutions rather than defer to traditional law. Modern man has replaced science for religion, and on the Net, the architecture of computer code becomes a substitute for jurisprudence.

However, Lessig errs when he includes "Internet governance" into this framework rather than treating it independently from the traditional areas of public policy that have migrated to the Net. "Internet governance" concerns the "subficial" aspects of the Internet -- the technical management of Internet infrastructure. It refers to how Internet service providers, protocol developers and network operators have over time voluntarily agreed to respect a central authority to set policy decisions over the technical underpinnings that enable the Internet to function.

The administrative tasks, performed for over two decades by the late Jon Postel of the Internet Assigned Numbers Authority (IANA), concern issues such as domain name system operations, Internet Protocol numbers used for routing the digital traffic, the "root servers" that link names to numbers, and the assignment of special networking parameters, such as port numbers. It's techy stuff.

These areas have recently become political because IANA's legitimacy to set policy was never established in law. Also, as a central point of control for the Internet, the oversight of these tasks represents a major source of power. Although the term "governance" is often used to describe IANA's clerical tasks, some, like Brian Carpenter, the chairman of the Internet Architecture Board, believe it is better described as "coordination" or "management." In fact, the term "Internet governance" in this sphere is actually an abbreviated form of saying "Internet self-governance" -- the way in which the ganglia of privately-owned interconnected networks that make up the Internet all come together to interoperate. It refers to how the Internet developed its own institutions to govern its own actions.

Yet this notion of "governance" is quite different compared to what is meant by "governance of the Internet" although it shares the same term. In the first instance, the term governance refers to regulation by government in customary areas of policy. In this instance, it refers to the process of self-organization for the technical coordination of the medium by the actors that embody it. The former denotes social choices by political leaders to serve their citizens; the latter marks the voluntary acceptance by autonomous networks to jointly manage themselves, since they individually represent the distributed resources of the Internet.

A failure to distinguish between these two areas leads to an inaccurate understanding of the dynamic behind the United States government's transfer of control of IANA's responsibilities to ICANN between January and November 1998.


While governments may be new at migrating the traditional areas of law onto the Net, the software developers, equipment manufacturers and network operators that comprise the physical transmission infrastructure itself have had a long and successful history of governing their own affairs. This history helps explain why the U.S. ceded its authority and funding of IANA to the private sector in an international and bottom-up organization that ICANN hopes to embody.

While the action was partly due to a typically American mistrust of government, it was made feasible because the Internet's governing mechanisms were already in place and worked well. They simply needed to be formalized to meet the future requirements of users, and made more representative and accountable on an international basis.

At the crux of the transfer of control from IANA to ICANN lies the question "What is the Internet?" For users, it is what they see in their browsers, the content they receive and the things that they do, like sending credit card information to buy a book online. The sorts of things that governments will regulate.

Bet there's another Internet that is not as well understood by the public, or by policy makers. It is the way the system works. There are four ingredients: the protocol, the names, the numbers and the so-called "root servers".

To start with, the Internet has a common language, called Transmission Control Protocol/Internet Protocol (TCP/IP), which all sites and all networks have to use to interact. It was created by two engineers in the 1970s working under a grant by the U.S. Department of Defense's Advanced Research Projects Agency, or DARPA. But hundreds of network engineers around the world built upon it, often voluntarily. TCP/IP, and all its enhancements became non-proprietary standards, which allowed anyone to freely implement it. Today, protocol development is coordinated by the Internet Engineering Task Force (IETF), the Net's standards body.

With a common protocol, networks are needed -- the actual telecommunications circuits that link users to "host" computers, like Web sites. Internet service providers carry the long-haul traffic, as well as supply connectivity to customers like corporations and individuals. These privately-owned, transnational ISPs agreed, for the sake of interoperability, to respect the authority of a central arbiter for networking decisions. They deferred to Postel at IANA, based at the University of Southern California, who was respected for being trustworthy and fair in setting policies that represented the consensus of the emerging Internet community of ISPs, sites, and protocol developers. IANA was formed in the late 1980s, but the functions Postel performed had begun in the mid-1970s.

Once a common protocol and networks exist, next come the Internet sites themselves. They require numeric addresses, called Internet Protocol numbers, to route the network traffic. IANA is the ultimate overseer of the finite IP numbers to ensure that they are handed out efficiently and not accidentally duplicated. But the day-to-day allocation of the resource is done by three regional registries: the Asia-Pacific Network Information Center (APNIC) in Brisbane, Australia; the Reseaux Internet Protocol Européens Network Coordination Center (RIPE NCC) in Amsterdam, the Netherlands; and the American Registry for Internet Numbers (ARIN) in Herndon, Virginia. All three institutions are non-profit, membership-based organizations that have evolved since the early 1990s. Their policies are determined by their members, the users of IP addresses -- and which come from nearly every country in the world.

Of course IP numbers are used by computers, but humans prefer words. In the early 1980s, the domain name system was invented to link the numbers to user-friendly names. The original domains, like .gov and .com, where quickly expanded to include country-code specific domains like .ch for Switzerland. IANA was responsible for assigning the rights to operate these country code registries to private individuals to serve the nation's Internet deployment. These individuals came from academia or the private sector, but rarely from government. When asked why he didn't simply delegate this responsibility to governments, via national post offices or telecom carriers, Postel replied that if he had, the Net might not have penetrated those countries as quickly since governments tended to act slowly and bureaucratically.

Finally, IANA oversaw the deployment of "root servers," the Net's addressing system that link the names to the numbers -- vital pieces of Internet infrastructure. Some root server operators work on behalf of the U.S. government, but most perform their tasks on a volunteer basis from academia or the non-profit or private sector.

While IANA received funding from the U.S. government via a convoluted series of contracts with the private sector, the government maintained a very hands-off approach. The system worked. IANA received its legitimacy from the global Internet community itself, like software companies, ISPs, and major corporate users. At any time, some or all of these players could have deviated from IANA's decisions, but none did since it was in everyone's larger interest to respect a single, central authority and remain connected to everyone else.

The process was called Internet self-governance. And an informal process existed for members of the Internet community to modify policies through the institutions that the Internet itself created to coordinate its operations.

As the Internet grew, the informal process would need to become more structured. It also had to include a greater swath of voices -- no longer simply the engineers who built the Net, but the people who used it and the companies that invested in it. However, the institutions already existed internationally to oversee the Net's development, which would serve as the framework for IANA's successor organization, ICANN, created in November 1998.


This notion of governance was always outside of government. Due to the Net's fast pace and need for flexibility -- which is at odds with traditional public policy -- governments today seem willing to let the Internet community manage its own technical affairs. ICANN has a Government Advisory Committee to represent the voice of nations, but it has no decision-making powers.

That's a good thing, since Internet engineers are loath to accept government authority, having built cyberspace outside of any government control -- other than light funding from the U.S. government for IANA's functions. Yet there is another reason why the U.S. government deferred to the Internet community rather than governmental or international bodies -- and lobbied other governments to follow suit: The history of the Internet is riddled with examples of "The Establishment" trying to kill off the Internet since it posed a major technical and commercial threat to the status quo.

In the 1980s, the International Standards Organization (ISO), formally backed the "Open Systems Interconnection" (OSI) protocol and refused to recognize TCP/IP. The Internet specification was not developed by a government-backed standards body, and so governments believed they could not control it as easily. This, despite the fact that OSI was only a theoretical system at the time, while the Internet was actually deployed. Many state-run carriers had invested in OSI, so lobbied their governments to support it.

The International Telecommunication Union (ITU) also refused to recognize Internet Protocol until the mid-1990s. It supported a costly and cumbersome data networking standard that their members, the state-run telecoms, developed, called X.400. The ITU also acted as a global registry for the X.400 addressing system, so had a vested interest in promoting its use. The U.S. Department of State actually prohibited the ITU from recognizing Internet standards since the IETF was not an official standards body. The ITU only agreed to officially accept Internet protocols as a part of its own standards-setting activity as recently as October 1998.

Telecom carriers were threatened by the Internet since it meant a loss of control over the usage of their network, and its users. In the early 1970s, AT&T initially refused to lease a nationwide circuit to BBN Corp. and research institutions for the Internet's precursor, ARPANET, until the U.S. government stepped in to force the carrier to do so. In France, the state-run carrier France Telecom was slow to develop Internet technologies in the mid-1990s because it wanted to preserve its lucrative but technologically inferior Minitel system.

These incidents demonstrate how both governments and corporate interests allied to government at times tried to sidetrack the Internet in its early stages in a bid to protect the status quo. It is a testament to the Net and the engineers who developed it that the Internet was able to surmount these institutional obstacles.


Cyberspace, according to pundits as diverse as the American poet John Perry Barlow or the former French minister Jacques Attali, is the "seventh continent"; the "New World" of the next millennium. And it follows nicely that to consider the Internet and governance, Professor Lessig recalls an earlier point in history -- the New World that North American once represented -- in citing the United States' 1803 purchase of the Louisiana territories from Napoleon. For the American national identity, it serves as a powerful symbol of the frontier, a place where rugged individualism reigned beyond the control of civil society's law; a sort of Rousseau-like "state of nature" where man was free and the master of his destiny.

But it's only one of a number of possible historical analogies. Although it suits Lessig's perspectives in the case of "governing the Internet" -- for the young nation's law tamed that land -- it is not the most appropriate analogy for "Internet governance". Rather, the land grants of the mid-1800s by the U.S. government to settlers in the Midwest and railroad companies serves as a better reference for the events from 1997 to 1999. Just as the U.S. sought to develop the vast territory to its fullest by giving away its land for private use, so too does U.S. policy today seek to deploy the Internet most widely by giving away its control to the international community. In this case, by letting the users of the resources manage its allocation.

So is it governance that is reviled. Or does the creative anarchy of the Internet naturally require an absence of traditional government institutions if it is to develop in an un-encumbered manner. Critical in the attempt to answer these questions is a complete understanding of the difference between governments controlling the usage of the Net, as opposed to the Net managing its own internal, technical affairs. To be sure, some of the technical issues have become larger policy questions, such as the trademark protection of domain names. Still, the distinction remains. And in the case of ICANN, the broader principle has been preserved -- not established, but preserved -- that the Internet community itself is responsible for its own development. Less a sign of disgust in government, ICANN might harbinger governmental obsolescence.

Indeed, an entirely new vocabulary of governance has emerged. Where kings once ruled subjects, and democratic governments represented citizens, the new social contract for cyberspace embodies "stakeholders" forging "consensus" to benefit "users." A new medium like the Internet may require new institutions that match its unique characteristics -- such as being informal, flexible and private. Governments are perfectly suited to support the interests of their citizens from which their legitimacy to govern is derived in a democracy, such as in regulating activities over the Internet. But in terms of coordinating the Net itself, it is not clear that the nation-state is the most appropriate mechanisms. It would be irresponsibly presumptuous to believe so simply out of nostalgia. In this instance, government might be the ancien regime.

If behind the Internet lies a new form of policy architecture that substitutes computer code for constitutional law, as Lessig suggests, than it becomes apparent that the Internet is not only an American technological invention but one that encompasses the nation's social values as well. Decentralization, individual empowerment, resilience and self-sufficiency are all design principles of the network as well as American cultural traits. Those values are now engineered into the network's governance structure with ICANN. Similarly, the Internet only fully emerged when it left academia for the private sector, and its governance is a reflection of that power shift.

As is the broader current of modern international relations. It too has had to evolve. The commercial world of multinational corporations today rivals the size and influence of national governments. New supranational institutions, such as the European Commission or the World Trade Organization, have emerged to match them. And to rule this new medium dominated by the commercial sector, a private international body, ICANN, has formed.

Ultimately, these broader trends of relying on the "logic of the marketplace" might not be such a good thing, but that is a question for another time. It is certainly not a very good thing for constitutional lawyers, like Lessig. Perhaps they should have specialized in contract law instead. There are jobs waiting in Hong Kong.


# # #


Copyright 1999 Kenneth Neil Cukier (