United States: Emerging Antitrust Issues In E-Commerce

The Internet is everywhere. Nineteen ninety-six was the year of the Internet. So was 1997. And 1998. And 1999. Year 2000 undoubtedly will be the year of the Internet, too. The Internet, it is said, is the "big equalizer." All one needs to set up shop on the Internet is a few hundred dollars for a server and a phone line. Entry is easy, amazingly so. The playing fields are level as they have never been before.

Will the Internet relegate antitrust to the trash heap for obsolete transistor radios and horse carriages? Or are there real pockets of potential market power—bottlenecks—for firms to find and exploit? Given the extraordinarily low barriers to entry and inherently atomistic nature of competition on the Internet, one can expect to hear impassioned arguments that the best direction for antitrust is to pack up and go home.

But the reality is different. There are many opportunities in e-commerce for the creation and exploitation of market power, and antitrust will play its traditional, crucial role in allowing the market to function freely and competitively without the need for regulatory control.

Four examples provide a useful perspective on the antitrust issues that have arisen and will continue to arise as e-commerce expands.

The first is distribution. Can serious antitrust issues really arise in the context of distribution of goods sold over the Internet? The answer is surely yes, as Barnes & Noble can attest.

Earlier this year, Barnes & Noble agreed in principle to acquire the Ingram Book Group. Barnes & Noble is one of the nation's leading booksellers, and (through a joint venture) the number two seller of books over the Internet behind Amazon.com^tm. Ingram is the largest book wholesaler in the country, and one of the primary sources of books for book retailers across the country. Given the importance of book wholesaling as a means of distribution in the industry, Amazon.com^tm and a number of independent booksellers protested the acquisition to the staff of the Federal Trade Commission (FTC), and the staff recommended a suit to block the deal. Immediately after announcement of the staff's opposition, Barnes & Noble abandoned the transaction.

Had Barnes & Noble acquired Ingram, at least in theory, it could have disadvantaged its rivals at the retail level by discriminating in price, access to hot sellers during short supply, returns policy, and other means, and by gaining access to rivals' confidential sales information. Again, in theory, this could have led to a raising of its rivals' costs, giving Barnes & Noble some power over price at the retail level. None of this would be possible if entry into book wholesaling was easy and sufficient, but it is not at all clear that it is.

The Barnes & Noble/Ingram transaction illustrates a key point about e-commerce: however easy it may be to open a retail Web site, Internet retailers still need access to actual goods and the means to deliver those goods to businesses and consumers. Bottlenecks may arise in distribution for e-commerce, therefore, in much the same way as they arise in the "brick-and-mortar" world. When they do, market power may be created or enhanced and antitrust will be needed to play its traditional role.

A second example is standard setting. As e-commerce grows, technological development has accelerated at an astonishing pace. For many technologies, however, utility depends on the ability of users to settle on a single standard for hardware and software.

An example is DVD. Manufacturers of software—such as DVD movies—will be able to compete much more efficiently if the product can be manufactured to one specification rather than two or 20. Having one specification is a boon to consumers, too, by eliminating the concern as to whether a particular film or concert disk will work in their home machines and by allowing achievement of economies of scale in software production. But settling on a single standard necessarily requires concerted action among horizontal competitors at the hardware manufacture level, raising potentially serious antitrust concerns. Collusion among competitors can lead to depression of innovation, increased prices, reduced output and quality and the exclusion of new entrants.

The Antitrust Division of the Department of Justice faced this precise issue twice in the last year. In December 1998 it approved a DVD patent pool arrangement among Phillips, Sony, and Pioneer, and in June 1999 it approved a similar DVD arrangement among Toshiba, Hitachi, Matsushita, Mitsubishi, TimeWarner, and JVC. The Department has approved a number of such arrangements over the last several years, including the MPEG-2 standard arrangement in June 1997. The business review letters approving these arrangements were all based on contexts in which the patents were complementary (rather than directly competing technologies), the patents were necessary to practice the technology, the pool resolved potential blocking conflicts, the parties agreed to license anyone on a non-discriminatory basis, and licenses to individual patents were available. These facts, frankly, make the issue an easy one. The analysis would be harder if the technologies were competitive -- such as Beta versus VHS -- or if licenses were restricted to the members. How the agencies and the courts will resolve these more difficult issues remains to be seen.

A third example of antitrust concerns that will continue to arise as e-commerce grows -- intellectual property issues -- permeates several of the others. Intellectual property rights inherently give rise to questions of potential market power, and the ways in which that power is exercised can often give rise to antitrust concerns.

One of the most interesting and widely followed cases in this respect was the FTC's proceeding against Intel, which was settled earlier this year. The case against Intel was unusual. The allegations were that, in the context of intellectual property disputes with Digital, Compaq and Intergraph, Intel withheld advance technical information and product samples. Normally, firms are fully authorized to take legal action to enforce their rights, including intellectual property rights, unless the legal action is "objectively baseless." There was no suggestion that Intel's IP disputes met that standard. Normally too, firms are allowed to withhold advance technical information and product samples from anyone, including those with whom an IP dispute is brewing. The FTC’s theory was that, notwithstanding these rights, Intel was using its monopoly power in microprocessors to coerce the three companies to license their microprocessor improvement patents back to Intel. The contrary argument -- questioning why Intel should have to provide special benefits to companies with which it has a dispute -- has some weight, but with Intel's agreement to a consent decree those issues remain unresolved.

A fourth example is the broad one of market power. As the preceding discussion has indicated, market power issues are not going to be eliminated by the growth of the Internet. To be sure, there will be changes. In many areas, barriers to entry will be lowered dramatically and, correspondingly, competition will become much more atomistic than before. At the same time, however, the importance of intellectual property rights will increase and, as that happens, the market power concerns traditionally raised in brick-and-mortar industrial contexts will be revived, though in slightly different forms.

Surely the best current exemplar of that phenomenon is the Microsoft case, where the core issue is the degree to which Microsoft's ownership of Windows -- fundamentally, an intellectual property right -- gives it the right to control the sale of other types of software applications by making them part of Windows or otherwise.

It is difficult to accept the notion that Microsoft has no monopoly power in personal computer operating systems. Virtually everyone today needs a personal computer. If a person wants a personal computer to work, the person needs an operating system. There are few choices of operating system and, if a person wants the one for which the vast majority of applications software is written, wants to share files and documents with colleagues, or wants a personal computer made with the fastest microprocessors, hard disks, and features, the person must have Windows. The idea that Linux, or Macintosh, or some future technology is sufficient to constrain Microsoft's existing power in selling operating systems is tough to swallow.

The more serious question is whether Microsoft's conduct has been competitively harmful. There is a very credible argument that the integration of Internet Explorer (IE) into Windows provides consumer benefits, and there certainly has been no obvious increase in price or restriction of output of Internet browsers from anything Microsoft has done. Moreover, consumers have ample access to Netscape; it can be downloaded and installed for free.

But the story may not be so simple. A few years ago, there was intense competition in browsers. Netscape 2.0 was a breakthrough, as was version 3.0, and the succeeding versions of both IE and Netscape advanced the technology to unprecedented levels. Indeed, that competition played no small role in the growth of the Internet itself.

Today there is Internet Explorer 5.0. It is a great product. But there is nothing like the advances that were taking place practically every week a few years ago. Will there ever be Netscape 5.0? Or is Netscape now just a portal for surfers to pause on while they move quickly to something else with their IE5? Has the pace of innovation slowed? The answers appear to be self-evident, and they compel the conclusion that there is a lot of substance to the government's case.

Competition on the Internet today is surely a lot different than competition in the past. But whatever the differences are in shape, manner and form, it may very well be necessary to monitor restraints on competition today with the same level of vigilance as in the past. Internet or no, antitrust is not going away any time soon.