Comment of the Staff of the Federal Trade Commission

Docket No. 9505 31 44-5144-01

The staff of the Federal Trade Commission(1) (“Commission”) appreciates this opportunity to offer comments on the Patent and Trademark Office’s (“PTO”) Proposed Examination Guidelines for Computer-Related Inventions (the “Guidelines”).(2) The Commission is an independent administrative agency charged by statute with preventing unfair methods of competition and unfair or deceptive acts or practices in or affecting commerce.(3) The Commission is one of the two federal agencies with principal responsibility for enforcing the antitrust laws.

Intellectual property law and antitrust law share the common goal of “encouraging innovation, industry and competition.(4) Thus, when properly applied, the two bodies of law complement and reinforce each other’s purposes. Conversely, inappropriate application of either can undermine the purpose of both. If antitrust enforcement unnecessarily prevents intellectual property owners from profiting from invention, this interference also may compromise the goals of antitrust laws.(5) And inappropriate or overbroad grants of intellectual property rights may interfere with the competition that often drives innovation.(6)

The intellectual property laws grant exclusive use of intellectual property to its inventor, designer, or author for a limited period of time, in exchange for disclosure of the invention. This grant is intended, in the words of the Constitution, “to promote the progress of science and useful arts” by enabling creators to claim the rewards of ownership. Intellectual property protection that is broader than is appropriate for the scope of the invention, however, may pose a threat to competition and result in higher prices to consumers.

Overly broad intellectual property protection also may reduce innovation by other inventors who fear infringing on the broadly patented interests. This risk is especially acute when the innovative process at issue is characterized by the accumulation of relatively small steps, rather than discrete leaps, and thus runs a greater risk of infringing possibly overbroad prior patents.(7) This type of innovative process is characteristic of software technology.(8) The impediments to future innovation created by inappropriate patents can be heightened by strong “network effects,(9) which are also characteristic of software. That is, new software will have greater value to the extent it is compatible with older systems and with the existing hardware and software base. If a patent is inappropriately granted to software that is already part of the prior art and has become embodied in existing software products, interfaces, and approaches, both the producers of current products and would-be innovators may find it very difficult to devise alternate technical solutions acceptable to the marketplace.(10)

The risk of harm from overbroad patents is, of course, not unique to the software industry, but a number of practical problems that have been identified by the PTO underline the need for caution about expanding the scope of patentable subject matter in the software area.

The subject matter test, the novelty test, and the nonobviousness test protect against overbroad patent protection for software. As the PTO has recognized, the novelty and nonobviousness tests do not function as well at this time in software as they do in other areas. Prior art can be difficult to identify in the software area.(11) The reasons have been stated by Commissioner Lehman:

[In] improving quality of examination for our software-related inventions … there are some inherent problems there which are very difficult to address … [including] that a lot of what is known in this area is in the area of trade secrecy. It’s not written down anyplace. It’s not even in prior patent applications, and so we have a very difficult time sometimes making determinations.(12)

The proposed Guidelines, however, would tend to reduce the importance of the threshold subject matter test, which has been the central issue in software patenting.(13) To the extent that the Guidelines go beyond what is required by In re Alappat(14) and subsequent cases,(15) caution is appropriate. Easing the subject matter test, as it applies to patent applications for software, apparently would mean greater reliance on the novelty and nonobviousness tests to weed out inappropriate patents. The PTO has recognized that there are currently practical problems in relying on the novelty and nonobviousness tests to allow patents for truly innovative ideas while rejecting the mundane.(16) Such practical problems, and the adverse consequences for competition and innovation that flow from inappropriately broad patents, underscore the need to examine carefully the basis for any changes that would have the effect of easing the subject matter test.

(1) These comments represent the views of the staff of the Bureaus of Competition and Economics of the Federal Trade Commission. They are not necessarily the views of the Federal Trade Commission or of any individual Commissioner. Inquiries regarding this comment should be directed to Willard K. Tom at (202) 326-2786 or Michael O. Wise at (202) 326-3344.

(2) 60 Fed. Reg. 28,778 (June 2, 1995).

(3) 15 U.S.C. § 45.

(4)” Atari Games Corp. v. Nintendo of America, Inc., 897 F.2d 1572, 1576 (Fed. Cir. 1990).

(5) This principle informs the Antitrust Guidelines for the Licensing of Intellectual Property, recently issued by the Commission and the Department of Justice. U. S. Department of Justice and the Federal Trade Commission, Antitrust Guidelines for the Licensing of Intellectual Property (April 6, 1995).

(6) See Porter, The Competitive Advantage of Nations (1990).

(7) For a discussion of the distinction between “cumulative” and “discrete” technological development, see Merges and Nelson, On the Complex Economics of Patent Scope, 90 Colum. L. Rev. 839, 890-94 (1990). This article contends that, in other industries characterized by cumulative technology, broad blocking patents have retarded innovation and industry growth.

(8) Nelson, Intellectual Property Protection for Cumulative Systems Technology, 94 Colum. L. Rev. 2674, 2675 (1994). See also Merges & Nelson, supra at 902-04.

(9)” “Network effects” are present when the value of using a network for any one user increases with the number of other persons using the same network. Examples include telephone systems and VCR formats.

(10)That is not to say that patents are inappropriate in industries characterized by strong network effects. The point, rather, is that network effects can multiply the adverse impact of inappropriate patent protection.

(11) “Much of the prior art that the agency needs to examine [in the Compton case], however, cannot be found in patents or scholarly publications, but is buried in the industry communications of software engineers and in computer code. Unfortunately, patent examiners are not equipped to search in the right places.” Kramer, Computer Novelty: Can Software be Patented?, Legal Times, January 24, 1994, S33. See also Kenneth W. Dam, Some Economic Considerations in the Intellectual Property Protection of Software, 24 J. Legal Studies 321, 369 (1995) (expressing concern that innovation may be dampened because the database and classification system are not yet adequate to ensure that patents will not be issued for “inventions” that fail those tests).

(12) Public Hearing on Use of the Patent System to Protect Software-Related Inventions, Before Bruce A. Lehman, Assistant Secretary of Commerce and Commissioner of Patents and Trademarks, United States Patent and Trademark Office 2 (Jan. 26-27, 1995) (hereinafter “Software Patent Hearings”).

(13) Until quite recently, it has been assumed that software was patentable only as an element of a traditionally patentable industrial process, where patentees sought to foreclose use of algorithms only in conjunction with all steps in a claimed process. Samuelson, Benson Revisited: The Case Against Patent Protection for Algorithms and Other Computer Program- Related Inventions. 39 Emory L. J. 1026, 1096 (1990). Hence, the software patent debate has turned largely on subject matter questions, and the attempt to distinguish pre-emptive claims on mathematical algorithms (which are not patentable) from innovative implementations of mathematical algorithms that do not pre-empt use of the algorithm itself (which are potentially patentable). Id.

The Guidelines reduce the role of this initial distinction by instructing patent examiners to presume that (1) programmed computers are statutory “machines”, (2) program encoded memory devices (such as floppy discs) are statutory “articles of manufacture”, and (3) a sequence of steps implemented by code is a statutory “process”, all potentially eligible for patent protection under § 101 of the Patent Act. (35 U.S.C.A. 10, hereinafter, the “Act”.) Using these presumptions, patent examiners are then instructed to review claims for adequate specificity under § 112 of the Act. Under the Guidelines, operations performed at the direction of software may serve as “specific acts” that correspond to an invention claimed using “means plus function” language. If the Act’s specificity requirements are met, examiners then determine whether a claimed invention meets the statutory requirements of novelty and nonobviousness.

(14) 33 F.3d 1526 (Fed. Cir. 1994).

(15) E.g., In re Warmerdam, 33 F.3d 1354 (Fed. Cir. 1994); In re Lowry, 32 F.3d 1579 (Fed. Cir. 1994); In re Trovato, 42 F.3d 1376 (Fed. Cir. 1994).

(16) See, e.g., Software Patent Hearings at 2, discussed supra, text at note 12. Indeed, such practical problems, as well as other considerations, have resulted in controversy over whether patents should be granted at all for software. See Software Patent Hearings, supra note 12; Newell, Response: The Models are Broken! The Models are Broken!, 47 U. Pitt. L. Rev. 1023, 1026 (1986); Samuelson, supra; Samuelson, Davis, Kapor and Reichman, A Manifesto Concerning the Legal Protection of Computer Programs, 94 Colum. L. Rev. 2308 (1994); Stern, Solving the Algorithm Conundrum: After 1994 In the Federal Circuit Patent Law Needs a Radical Algorithmectomy, 22 AILPA Q. J. 170 (Spring 1994). But see Chisum, The Patentability of Algorithms, 47 U. Pitt. L. Rev. 959 (1986); Warren-Boulton, Baseman, and Woroch, Economies of Intellectual Property Protection for Software: The Proper Role for Copyright, 3 StandardView 1 (1995).