The Internet Policy Paradox: Less is More Except for policies that sustain critical functions, we should resist the urge to encumber the Internet with policies By Charles Brownstein cbrownst@cnri.reston.va.us Perhaps because of the scale of public investment in dot-com, or perhaps because of Alan Greenspan's recent notice of the Internet as a signal factor in the U.S. economy, Internet policy stock has soared. Academics and intergovernmental organizations confer on it. Institutes devoted to it spring forth whole cloth. Politicians grasp for it as a strategic magnet for attention (and, perhaps, for money). "e-(u-name-it)" is the new icon for social and economic progress, and hence it is elevated to the lofty heights of public policy. Otherwise reasonable people want to apply every possible remedy for the problems that ail society, and public policy is a tool for that. So it's increasingly common that poverty, literacy (and other aspects of education), economic development, health, privacy, antitrust, freedom of speech, social safety, national security, old age, and more are associated with Internet policy. Different people and different societies have vastly different perspectives. Some tend to see policy as the framework for action-to be fully developed before action can proceed-as in the European Commission. Others tend to see policy in terms of pragmatics-the rules of the road, often to be invented after the accidents are realized-as in the U.S. In either case, it seems to me that faith in public policy should be limited to areas in which it can have at least a chance of efficacy. Policy proponents must face the simple fact that the problems of humankind-whether societal or economic-require solutions of appropriate scale in size, interactivity, and time. The only hope for Internet policy is to ape the design of the Net. In some ways, the Internet is pure architecture-a high-level blueprint of a strategy to move bits among information processors over networks. That strong concept has been translated into an extraordinarily powerful implementation. Today's Internet keeps the relatively simple "bitway" functions at the center, and the more complex information processing functions-as well as the huge cost they entail-at the edges. That works well. It scales, and it is readily extended in ways that no one could have imagined beforehand. Costs are kept near the users, where markets can generate efficiencies. As a result, the function, performance, use, and impact of the Net just keep on growing. How, then, might one think of public policy in such terms? First, consider how to deal with the size of the problems one hopes to affect with the Internet. Take the digital divide. It was perhaps at its deepest back when the U.S. government was funding the ARPAnet, the NSFnet, and other research networks. It has been diminished only as advances in technology-many due to public R&D expenditures-have helped drive down costs for every technology involved in the Internet, from the optical devices that illuminate the Net to the software that constitutes its engines, to the ever-cheaper appliances that facilitate mass access. We are only a year or two away from premature worries about women and the Net, and we are just now understanding how readily seniors use the Net: to track stock portfolios, to e-mail their grandchildren, to access medical information (whatever its quality). Face it: using the Net is far easier than baking bread, and the first generation of software engineers is eligible for senior discounts. Or consider interactivity. Being able, yet not being required, to link everything is an element in the Net's multiple impacts. Think of the terms interdisciplinary, interagency, interinstitutional, international, and intercompany; or cross-discipline, cross-agency, cross-institutional, cross-national, and cross-industry. If you don't appreciate what the foregoing lists mean, you don't yet get it. The Internet is inherently a cooperative enterprise. Unlike platforms-such as operating systems and applications software-the Internet's value is determined ultimately not by what can be controlled and appropriated but by what can be shared and extended. When it is shared, there exists a space to sell goods and services and generate value and wealth. The information technology industry worldwide has learned the practical utility of "coopetition." Governments too often have not learned it. They continue to measure themselves against other governments in order to justify their Internet policies-either to promote economic competition or, less charitably, to justify control of content and society. But the Internet is a worldwide enterprise and phenomenon. "Internet is for everyone" captures the sense of network externalities. Institutions that try to appropriate regional or political advantages are doomed to self-marginalization. Finally, consider the effects of time. Public programs and overly detailed public policies cannot keep up with technological changes if they presuppose the details about what people want, need, can use, or will do. When tied to rapidly advancing technologies, applications research and policies intended for social equity must make assumptions about cost and performance that stretch the imagination beyond reasoned public policy. Such thinking in the future requires giving up past measures, no matter how hardwired they are in paralyzed government and commercial bureaucracies. Patience is the bane of politicians concerned with the next electoral adventure. Public policy is not likely to succeed if it is too detailed-or too intrusive. Where should Internet policy be focused? If, via public policy, we can drive down in every dimension the costs of the technology that composes the Internet, as well as the costs of access to it, then perhaps the Internet can be used to deal with the big problems. Lowering costs already has had a far greater effect than have subsidies for computers or networks or for their use. Internet "governance"? Contain it before it spreads-except for policies that sustain the critical function of coordinated numbering and naming and that prevent anything that could stifle competition and innovation. What about monetary and fiscal stability, the quality and content of education, personal privacy, economic development, health care, the environment, better user interfaces, protection of intellectual property, crime prevention, and preservation of economic competition in the face of monopoly power? These things are hard to achieve; they deserve maximal policy attention and maximal enforcement of the laws associated with them. We shouldn't encumber the Internet with policies about them, though. The Internet is far too important for what it can enable.