World Wide Grab?

By Ed Cohen | Spring 2012

Game theorists see mixed outcomes from abandoning a founding tenet of the Internet


As a computer user, you pay an Internet service provider or ISP for access to the Internet, and you pay extra for faster speeds. But website owners don’t pay ISPs anything to deliver their content to you. Why not?

That’s what ISPs such as Comcast, Time Warner and AT&T would like to know. And change.

There would be pluses and minuses to letting them, according to a study by three information-technology researchers, including Hong Guo of Mendoza College’s IT management faculty.

The researchers say the extra revenue in ISPs’ pockets could lead them to offer cheaper, even free Internet service to consumers. About 26 million Americans live in areas lacking broadband access, according to a 2011 Federal Communications Commission report.

But the IT researchers also warn that letting ISPs charge to carry content would likely depress website creation, and it would reduce ISPs’ incentive to improve speed and reliability.

As Guo and company explain in a recent journal article, the principal obstacle to ISPs making money off of websites is an egalitarian principle called net neutrality.

Dating to the Internet’s founding, net neutrality holds that no government or ISP should be allowed to restrict access to the Internet. People should be able to connect any way they like and consume as much bandwidth as they like. Net neutrality also mandates that every website compete on an equal footing when it comes to sending packets of data over the ever-more-crowded information super highway.

The ISPs want to offer a premium service to website owners. For a fee, they would guarantee faster transmissions with fewer interruptions over the so-called “last mile” of the Internet. That’s the piece the ISPs control, from the service provider’s local office to customers’ homes or offices.

To test the consequences of allowing ISPs to do that, Guo and her fellow researchers analyzed all possible scenarios through game theory. That’s a way of mathematically modeling conflict and cooperation among parties. It assumes that all parties will act rationally.

Their analysis found plenty of advantages to abandoning the equal-speed provisions of net neutrality.

Profits for ISPs increase. Internet service rates for consumers drop—under some scenarios to free—because the ISPs would have extra revenue from websites to subsidize rates and an incentive to deliver the largest possible audience to their website clients. Broadband market coverage increases as a result of the dropped Internet service rates.

Another advantage to offering faster, more reliable Web connections at a premium is it would enable services to develop that are extremely sensitive to interruptions, such as remote-controlled surgery.

The losers in the scenarios include many websites. Those that refuse to pay for premium transmission service would be at a competitive disadvantage because their sites might not function as well for consumers. Higher fees could put smaller operators out of business, reducing competition. High fees could also discourage startups, the researchers say.

ISPs have argued that allowing them to collect fees from websites would give them the resources they need to upgrade their infrastructure, which could help fix current Internet problems such as interruptions and slowdowns when playing online games or streaming video.

But the researchers’ analysis says that’s not likely to happen. Here’s why:

The current congestion on ISP lines is analogous to traffic on a crowded highway, Guo says. If people could pay a fee to jump ahead of other cars, many would. That would give the highway authority extra revenue. But the authority would then have no incentive to spend that money on adding more lanes. If it did, the traffic would go away and people would stop paying the jump-ahead fee.

“A similar argument applies to the ISPs,” says Guo, an assistant professor of management. “If ISPs can generate extra revenue from websites, then they aren’t going to expand capacity, because if they expand, the websites will not pay them for extra speed because the speed is already good.”

The fate of net neutrality remains unclear, at least in the United States, as courts, Congress and the Federal Communications Commission continue to debate the principle. The researchers say their analysis provides all parties insights into the likely consequences of changing the status quo.

The research team’s article, “Net Neutrality, Broadband Market Coverage, and Innovation at the Edge,” by Guo with Hsing Kenneth Cheng and Subhajyoti Bandyopadhyay of the University of Florida’s Warrington College of Business Administration, appears in the February 2012 issue of Decision Sciences.