The ongoing controversy over intellectual property protection for computer programs is the latest battle in the centuries-old war between those who favor intellectual property protection and those who oppose it. The conflict over computer programs, however, has an internecine aspect-if you will, a civil war among protectionists. It is not so much a question of whether computer programs should be protected, but rather which form of protection should be the exclusive or even dominant one. These combatants may be loosely grouped into various camps: pro-copyright, pro-patent, pro-sui generis (a separate title of protection), and those favoring or opposing various combinations of the foregoing. There are, of course, some anti-protectionists who would favor little, if any, intellectual protection for computer programs.

Under the current state of the law, it can be said with a reasonable degree of legal certainty that computer programs (including algorithms) qualify as the subject matter of both copyrights and patents, with perhaps a caveat in copyright law to account for the possibility of the merger of idea and expression and a caveat in patent law to account for ineptness or overzealousness in claiming. Starting from the premise of overlapping protectability, this article will analyze in a comparative manner the economic case in Part II and the general (legal and policy) case in Part III for the respective systems of copyright and patent protection for computer programs (including algorithms). In Part II, a model will be introduced for evaluating the costs and benefits of the respective systems. This model will consider whether the current systems of patent or copyright production induce a high yield of inventions or works of authorship directed to computer programs that would not otherwise be created but for the respective protective systems. In addition, the present systems will be evaluated to determine whether the incentives they provide are excessive, unnecessary, and costly in view of other, particularly market, incentives for the creation of computer programs. Next, various models of the interfaces of the patent and copyright systems as they relate to computer programs will be analyzed, with particular attention paid to the economic consequences that may result from a particular legal determination of the interface model.

In Part III, legal and policy considerations will be comparatively addressed as they relate to the general case for or against patent protection for computer programs, including: the anti-patent sentiment; protection of processes not involving the transformation of matter, including methods of doing business, in an evolving service economy; the specter of patent infringement by thinking; the constitutionality of protecting programs by both copyright and patent; assumptions concerning a causal relation between intellectual property and the growth of the computer industry; and advocacy for and problems associated with the establishment of a sui generis system of protection for computer programs.

The conclusions drawn, on a relative basis, are that a somewhat easier (but not an entirely easy) overall case can be made for patent protection of computer programs than for copyright protection under the current state of patent and copyright law and that a sui generis system offers speculative advantages, at best, over the present systems.