COMMENTS OF W. RUSSELL WAYMAN,
CORPORATE VICE PRESIDENT AND GENERAL COUNSEL OF
STORAGE TECHNOLOGY CORPORATION BEFORE THE
FEDERAL TRADE COMMISSION

NOVEMBER 30, 1995

Thank you for the opportunity to speak with you this morning. As the General Counsel of Storage Technology corporation, I provide general legal advice and supervision of the legal affairs of a $2 billion dollar company engaged in the development, manufacture and sale of computer memory and networking products. These subsystems are principally used in large computer system environments. I have been a lawyer in high tech industries for over fifteen years.

I think it is vitally important that the government focus on high technology. The high tech sector of U.S. business has been a very successful source of employment and export revenue for our country. The health of this industry is as important to the U.S. as that of any of our key commercial enterprises. I further believe that antitrust law is an important component of the governmental stewardship necessary to assure the continued good health of this industry. I do not believe that this industry will flourish without the help of antitrust law, and I do not believe that this industry will be handicapped vis-a-vis its foreign competition by the enforcement of our antitrust laws.

I have been asked to give you some thoughts as to how computer companies compete. What follows are just that, a few thoughts. I hope they will be somewhat illuminating, but I can assure you they are not offered as everything you need to know about high tech and antitrust. I want to give you three somewhat distinct thoughts: first, two key differences I see between high tech companies and more traditional companies; second, the peculiar nature of software in intellectual property law; and third, the importance of interfaces.

Two Key Differences

In many ways, computer companies, or other high tech organizations, compete with the same strategies and goals of those of any other commercial enterprise. There are, however, areas of unusual emphasis in the computer industry which deserve special acknowledgment. I want to focus this morning on two key differences between traditional industries and the computer industry. These special characteristics have important implications for enforcement of antitrust law in high tech industries. As I look at competition in my industry, I am first struck by both the pace of technological change and, second, by the importance of intellectual property.

The rapid pace of technological change in high tech industries is probably universally appreciated. All of us have watched the price and functionality of the personal computer change rapidly over the last few years. These advancements have been in large part driven by similar changes in speed and capacity of the various components, most particularly microprocessors and computer disk drives. In fact, virtually everywhere one looks in high tech, one sees examples of rapid advancement of the technical frontiers.

In addition to rapidity of change, I believe a second characteristic of our industry which distinguishes it from traditional industries is the importance of intellectual property. While intellectual property is part of many companies' asset base, it is more important to high tech companies. If one were to compare, say, the old United States Steel Corporation with Microsoft, and look for the most valuable assets of each, the comparison would, I believe, yield striking dissimilarities. The assets of a steel mill are largely found in the immense capital structures necessary for these enterprises to produce their products. The blast furnaces, rolling mills, and other steelmaking equipment are the backbone of such an enterprise. Certainly the capital structures of a software company are about as different from a steel company as could be imagined. The principal assets of Microsoft Corporation, or any successful software company, are the programmers that develop the software. However, unlike steel mills, those programmers are not assets which Microsoft "owns" in any traditional sense. The owned asset is the intellectual property which those programmers create. In some sense, the entire multibillion dollar value of Microsoft Corporation could walk out the door some day in the same way as Michael Jordan might depart from the Chicago Bulls. In another sense, however, Microsoft would be left with the work that those programmers had done. It is obviously imperative to Microsoft stockholders that programmers' work be protected from use by companies which did not pay to develop the software and which do not have the legal right to use it. Such protection is afforded by our intellectual property laws.

It is a peculiar attribute of intellectual property that it is often infinitely reproducible at virtually no cost. In the case of a steel mill, even if there were no legal barriers to replication of the plant, the cost of replication alone would constitute a significant barrier to competition. In the case of a software company, if there were no legal barriers to copying, any competitor would be able to replicate that enterprise, at least in the sense of looking backwards, at virtually no cost. Does this mean that intellectual property law is the only protection a high tech company has to guard its asset base? I would submit that the answer is no; high tech companies have one other difference from traditional companies which can protect the economic benefit of their intellectual property. Even if there were no intellectual property laws, and a third party was able to completely replicate the software repertoire of, for example, Microsoft with legal impunity, Microsoft would still maintain a significant competitive advantage because of the other peculiar characteristic of high tech industries, that is the rapidity of change. By way of analogy, the ability to completely replicate a race car will be of extremely limited value if your race car is parked at the side of the track while the competitor's race car is moving down the straightaway at 200 miles an hour. The fact that you had something that exactly equaled your competitor at the initial instant of its replication is almost totally irrelevant as your competitor speeds away.

Copying of intellectual property of a software company is very much analogous to replicating a race car, but in a standing still configuration. I would, therefore, submit that the competitive advantage of a computer company or other high tech company is really based on two characteristics, the ability to protect its intellectual property in a legal regime, and the rapidity with which it can augment and extend that intellectual property.

Consider now the public policy implications of these characteristics. A corporation in a high tech industry has two ways of sustaining its competitive advantage. The first is to erect a very high barrier of intellectual property protection -- the higher the barrier, the less quickly it must run. That is to say, if there was no intellectual property protection, the only significant protection would be to continue to move so rapidly to introduce new, faster and less expensive products that no competitor could keep up, even if that competitor copied the characteristics of the company at one particular point in time. The other manner of operation would be to remain relatively static but rely on intellectual property protection to shield the corporation from the ravages of competitors. Let's compare those two ways of competing from the standpoint of consumer welfare. I think the answer is obvious. Clearly, the consumer is best served by encouraging a regime within which the best defense of any company is to attempt to run faster than any of its competitors. The result of such a corporate paradigm is newer, faster, better, cheaper products on a regular basis. To the extent that a company eschews this competitive model and chooses instead to rely on an entrenched intellectual property position, the result is an ossified marketplace in which the first movers competitive advantage is sustained by its legal department and the courts, as opposed to the rapidity with which it introduces new ideas and products. In other words, in the high tech industries the common base assumption in evaluating intellectual property may not be true. Companies may be induced to do more R&D and introduce new products faster if it is relatively more difficult to protect intellectual property.

How should such a view of the computer industry affect the policies of the antitrust law? Simply put, the most important thing that the antitrust law must do is to focus on the use and abuse of intellectual property rights by high tech companies. The existence of intellectual property (patents, trademarks, copyrights, and trade secrets) is well established in this nation, and I don't believe that we would be well served by the dismantling of that regime. However, it is subject to abuse by firms holding dominant positions within the high tech industry.

The Peculiar Nature of Software

I believe that it is at the margins of established intellectual property law that the antitrust law should look for attempts to extend those rights beyond their historically appropriate confines. Those boundaries are most unclear with regard to software. Software, or firmware, or the logic structure of a computer is a relatively new phenomenon. The intellectual property bar has been struggling with software for many years now. It was recognized relatively early on that software was a creature which did not clearly fit within the traditional patterns. The CONTU report was a seminal effort by the legal establishment, with significant assistance from its brethren in the computer sciences field, to understand software and to place it in an appropriate context within the legal regime.

If I compare the intellectual property regime which has evolved for software with the regime for other kinds of property, I think the contrast is quite striking. In the beginning, it was not clear that software was protectible under federal intellectual property law. In fact, the courts had indicated in early decisions that copyright might not apply to software. See Data Cash Sys. Inc. v. JS&A Group, Inc., 480 F. Supp. 1063 (N.D. Ill. 1979), aff'd on other grounds, 628 F.2d 1038 (7th Cir. 1980). The courts had also indicated that patent law was not an appropriate vehicle for protection of software, Gottschalk v. Benson, 407 U.S. 63 (1972). The legal community, and I was one of the participants in this dialogue, had no clear way to protect its clients' significant investment. Understandably, we undertook to push for protection on every possible front. Since those early days, we have evolved an environment in which software is clearly protectible by patent, by copyright, and by trade secret. Compare that environment with the environment that is applicable to other forms of property. Protecting the intellectual property contained in an automobile is generally thought to be largely a matter of patent law. In contrast, for those persons who make a living writing songs or books, they look to the copyright law for protection. For industrial formulae, the good legal advice is that one may protect them by either patenting them, as is the case often in the pharmaceutical business, or by retaining those formulae as trade secrets. The Coca Cola formula is a time-worn example of such a trade secret based strategy. While one can understand the historical reason for the "triple threat" available for software protection, the contrast with other types of industrial products is striking. Essentially, because of the evolution of the law in dealing with software, we have a regime within which the software creators, and their counsel, purport to be able to sue their competitors for patent infringement, copyright infringement and trade secret infringement for the same piece of code.

Let me make it clear, my company regularly creates vast amounts of software. My client, like other competitors in this industry, would be devastated if there were no clear way to protect its software. Nevertheless, I think this peculiar triple threat is a matter for some reflection by regulators in looking at the overall high tech landscape and the importance of an appropriate balance between encouraging creation of intellectual property by protecting it, and encouraging competition by necessitating continued development of new and better products, that is, running fast.

The Importance of Interfaces

Let me conclude with a few specific remarks as to my perception as to how computer companies compete. Remember, I'm not suggesting that we throw overboard the many years of effort we have collectively put forth in attempting to more clearly define the appropriate scope of intellectual property protection for software. I am, however, suggesting that, at the margins, we ought to be conservative about extension of those rights. Where are the margins? The margins are the point at which software of one company interfaces with software of another company. In these areas, we ought to facilitate attachments. We ought to break down the intellectual property barriers which frustrate them. Let me give what I think is a very apt metaphor and then let me analogize with some antitrust cases.

General Motors makes automobile engines and carburetors. The connection between an automobile engine and a carburetor is effected by bolting the carburetor to the engine's intake manifold. In order to make that attachment, it is necessary that the carburetor be designed to exactly accept the bolts extending out of the intake manifold. The interface between the carburetor and manifold must be exactly replicated in order for the carburetor to effectively function. There are other companies besides General Motors which wish to make carburetors for attachment to General Motors' engines. They wish to compete in the market for carburetors. In order to do that, they need to disassemble a General Motors engine, discern the pattern of the bolts and replicate that bolt pattern, in order to attach their new carburetor to the manifold. To the extent that the intellectual property laws frustrate the ability of the carburetor manufacturer to look at the bolt pattern to develop a compatible interface, the laws frustrate competition and the consumer's choice.

In the software industry, the interaction between one piece of software and another, or between software and hardware, is much more subtle and interactive than a mere bolt pattern. Accordingly, on occasion it may meet a standard which makes the interface patentable. At the same time, given to critical role of interfaces, we should be wary of extending copyright, with its low threshold of originality to such interfaces.

Antitrust law has regularly struggled with enterprises that control interfaces. Some of these cases are found under the rubric of refusals to deal. Other cases are found under the theory of essential facilities. I am going to refrain from lecturing the Federal Trade Commission on antitrust laws, because I think that would be very much a matter of the student lecturing the teacher. You already understand the importance of this issue. The Commission's complaint against Silicon Graphics with respect to access to interfaces by independent software vendors, evidences your agency's awareness of the importance of the point that I am making.

Having said I will not lecture you on antitrust law, I do want to make one observation about these interface situations. I sometimes find a tremendous opposition to my beliefs in this regard by other computer vendors who perhaps wish to move more slowly and rely more on their legal departments, than their engineers, to protect their position. They point out the enormous investment made by their clients in software. But I am talking here not about that admittedly enormous investment in toto. I am talking about the interface between that investment and independently created pieces of software or hardware. This interface is not the principal value of the developed software. The enterprise wishing to develop and manufacture a connecting device (whether it is hardware or software) may not replicate the software of the first developer, only the interface between the first developer's product and the newly developed software. An appropriate metaphor is a bolt pattern. This should be distinguished from these cases in which the plaintiff was claiming that it must have access to the defendant's football stadium or to the defendant's railroad bridge in order to avail itself of an additional market opportunity. In those cases, the essential facility was itself a huge capital investment by the alleged monopolist. I submit that in the case of software interface specifications, we are dealing with an entirely different kind of animal. We are dealing with bolt hole patterns, not the right to use the competitors carburetor or engine. We are dealing with the right to copy the gauge of the railroad track, not the right to use the other party's tracks. And, if my tracks have the same gauge as you your tracks, everyone's train can roll seamlessly between the systems.

In conclusion, my thoughts are very much the comments of a participant in high tech, not an expert in antitrust or intellectual property law. But, hopefully, my impressions, that speed in innovation should be encouraged, that software is one of the most overprotected forms of intellectual property, and that open interfaces are crucial, will be of use to you in your future considerations.

I want to thank you for your time this morning. I very much appreciate your concerns with antitrust law in our industry. I hope that the Federal Trade Commission, along with the courts and the Justice Department, continue to believe that vigorous enforcement of antitrust policy is an essential component of preserving high tech industries global competitiveness.