Ask the Expert

Professor Ramesh Johari says: Informally, “network neutrality” refers to the idea that the owners of Internet infrastructure (companies such as Sprint or AT&T) should not regard different kinds of uses or users differently: Video downloads, for example, shouldn’t be billed at a higher rate than e-mail, and searches on one search site shouldn’t be billed at a different rate than searches on another (the term is thought to have been coined by law professor Tim Wu at Columbia University.).

This somewhat vague principle may sound like a cyber version of basic egalitarian fairness, but it is controversial because many people are unsure that a completely neutral network can be economically sustainable. Network users want freedom and innovation, but network builders need to recoup their investments in things like laying miles of fiber optic cable, deploying thousands of routers and switches, and building out wireless base stations.

Ed Whitacre, the former CEO of what was then called SBC, laid this issue bare years ago when he told Business Week that it would be “nuts” for companies such as Google and Yahoo! to think they could make billions using SBC’s communication lines without compensating the company. He was frustrated that while SBC (and other telecommunications companies, or telcos) invested heavily in providing the connection between users and those Web sites, the sites were the ones making all the money. Many Internet activists, however, bristled at the suggestion that telcos might start adding surcharges for valuable online activities.

Fundamentally the challenge, then, is to find a combination of technological and financial innovations that will make it sufficiently profitable to participate in building, maintaining and expanding an open, free-wheeling Internet. If struggling telcos tried to make the Internet closed and tightly controlled, like the old America Online or CompuServe networks, innovation and user experience would likely suffer a great deal. I find research in this area particularly exciting, because this is a problem that demands both a technological and economic solution; such research is emblematic of the Department of Management Science and Engineering. I am among many researchers around the world looking for ways that might keep the Internet free and open, but also sufficiently profitable for builders to keep building.

One possibility that has been suggested by some researchers is to try to infuse some of the principles of futures markets into the contracts that telcos and internet service providers agree upon when building out their networks. The basic principle is that in making a large investment in network capacity, the telco can simultaneously sell a futures contract for that capacity. A variety of mechanisms have been suggested to balance the later use of that capacity against the initial contracts; such mechanisms allow for a transfer of future benefit to the present to enable capital investment. Interesting questions arise regarding the optimal design of such contracts, and the construction of markets in which to trade such contracts.

I’ll conclude by noting the significant mismatch between data traffic and value on the Internet. All the contracts regarding Internet infrastructure are based on traffic alone, without regard to its economic value. For example, videos and peer-to-peer file sharing produce tremendous traffic but very little value. Credit card transactions and ads (like Google’s Adwords) produce tremendous value, with very little traffic. So the question is whether traffic alone is really the right basis for compensation. To more effectively distribute revenue among Internet users and Internet providers, shouldn’t the contracts in the middle account for value, instead of just traffic? I’m therefore investigating what mix of contracts and technological protocols might allow this value transfer to occur.

While my research is still in its early stages, we can already safely conclude that the answer will lie somewhere between the ideal but mythical “anything goes” neutral network, in which there are no distinctions at all among traffic, and the most extreme forms of locked down control where service providers determine what sites we can see, or what files we can download, based on who is paying them the most.

.