February 17, 1998

WOULD INTERNET MEGA-MERGER CHANGE THE LANDSCAPE OR JUST ACCENTUATE IT?

by Andy Oram
American Reporter Correspondent

CAMBRIDGE, MASS.—One wouldn’t think that any company could monopolize the Internet, this fast-growing and technologically exploding industry that defies anyone’s attempts to get his hands around it. But many are worried that a critical change could be triggered by a single merger: the proposed combination of WorldCom and MCI, two major providers of long-distance phone service and Internet access.

WorldCom, the country’s fourth-largest long-distance company, surprised the world on October 1, 1997 by snatching MCI away from an imminent merger with BT (also known as British Telecom) through offering their own bid of 30 billion dollars, later raised to 36.5 billion. Spurning both British Telecom and late-comer GTE, MCI accepted the deal on November 10.

But the merger cannot go through until approved by the Department of Justice and the FCC. Several parties have filed objections, including the Communications Workers of America (the major union in the telecom industry), the BellSouth Corporation (a telecom carrier), and the Consumer Project on Technology (a public-interest group started by Ralph Nader). Other long-distance companies have not commented.

Ultimately, I think the Internet is too fluid to be dominated. In fact, if the MCI-WorldCom merger goes through, it may end up being blamed for changes that recent developments hint may have to take place anyway. (As for Microsoft, it is so far from being able to dominate the Internet that it’s not even on the horizon.) But letting any one company get as big as the proposed MCI-WorldCom corporation—with up to 60% of all Internet traffic—leads to some risk.

Most of the revenue for both WorldCom and MCI comes from long-distance telephone service. In fact, shortly after proposing the merger, WorldCom chairperson Bernard Ebbers hinted that telephones would take up so much of their energy that they wouldn’t be able to think much about the Internet for a while.

But now the press releases from WorldCom indicate that they are very interested in expanding and upgrading Internet business. Having consummated the purchase of CompuServe two weeks ago, as well as the network portion of America Online, Ebbers shows that he doesn’t tend to do things incrementally.

Why are Internet users worried about the prospect of MCI-WorldCom? Well, start with a sudden policy change instituted last year by UUNET, the largest (and one of the oldest) Internet service providers, now a subsidiary of WorldCom. The policy concerns peering, a relatively subtle concept that deserves elucidation.

The whole notion of an Internet is predicated on the ability of different networks to exchange data. When the Internet was young, most traffic went over central wires maintained by a grant from the National Science Foundation. You can visualize the system as hundreds of small end-users attaching to larger branches, which then attach to still larger ones, all finally connected by one huge set of cables like the spine of a vertebrate—which is why the major cables are still called the “backbone.”

As private companies took over more of the backbone, they continued to let smaller companies exchange traffic over these cables and servers for free. This arrangement meant that large providers were essentially paying for equipment used by smaller ones. But the expenditure was worthwhile, because it meant more users could get on the Internet and ultimately everybody got more business.

Well, if you’re a Neanderthal and antelope are plentiful, you won’t mind fellow humans hunting the same hills, and may even give them a hand catching new game once in a while. But when the hills start to fill up with fellow Neanderthals, and Cro-Magnon man shows up, you get a good deal more territorial.

I don’t want to take this analogy too far, because I do not think by any means that the old Internet providers will die out, even when facing competition by cable TV and telephone companies. My point is simply that a maturing Internet industry makes UUNET feel the need to compete more with the small providers, as well as charge more for access to its servers and lines.

Whatever the motivation, UUNET has recently required small companies to pay more than ones that can offer facilities of equal capacity. Many people are afraid that this move signals a tug on the strings by WorldCom to force small providers out of business.

More harbingers of change are popping up. Traditionally, for instance, the Internet was dominated by flat-rate pricing, where you could stay online as long as you want—a boon to exploration and experimentation. But in the past week or two, a couple of Internet providers have introduced variations in the scheme: either two tiers of pricing for different quantities of usage, or per-hour charges for excessive use.

Combine this hint of a new trend with predictions that Internet providers must soon start providing different levels of “quality of service” (to support massive bandwidth use like video teleconferencing) and it’s easy to predict that flat-rate pricing will disappear.

But let’s not jump the gun: the marketplace has loudly stated its preference for flat-rate pricing, forcing mega-provider America Online to institute such a policy. (The recent announcement of a $2-per-month price hike at AOL does not affect the basic policy.) And most people will probably continue to depend on relatively low-bandwidth applications like electronic mail and viewing text or graphics on the Web. Flat-rate pricing suits them just fine.

A related danger lurks, though. The people using fancy applications like video teleconferencing are large institutions, particularly corporations that want to provide remote training. It’s much easier for a single Internet service provider to guarantee a certain quality of service on its own network than for multiple service providers to offer such a guarantee on a cooperative basis. The same goes for virtual private networking: the guarantee of secure connections that malicious intruders can’t snoop on or corrupt. And for multicasting, a technology that lets a single source send great volumes of data to multiple recipients.

In other words, more and more business reasons are emerging for choosing a single Internet service provider to serve all sites. Cooperating businesses will want to hook up with a single provider for the same reasons. If MCI-WorldCom really does end up controlling 60% of all traffic, they’ll have a powerful lever over their competitors.

MCI-Worldcom may prove to be a good Internet citizen. Internet-based companies have long been proud of conducting business with minimal regulation. But we have to watch out in case the “regulation” comes from a powerful monopoly rather than from an elected government.


Editor, O’Reilly Media
Author’s home page
Other articles in chronological order
Index to other articles