The patent statute creates a general set of legal rules that govern a wide variety of technologies. With only a few exceptions, the statute does not distinguish between different technologies in setting and applying legal standards. In theory, then, we have a uniform patent system that provides technology-neutral protection to all kinds of innovation. Technology, however, is anything but uniform, and displays highly diverse characteristics across different sectors. A wealth of empirical evidence demonstrates deep structural differences in how industries innovate. Industries vary in the speed and cost of Research and Development (“R&D”), in the ease with which inventions can be imitated by others, in the need for cumulative or interoperative innovation rather than stand-alone development, and in the extent to which patents cover entire products or merely components of products. We show that there is no reason to assume that a unitary patent system will optimally encourage innovation in the wide range of diverse industries that it is expected to cover.

Despite the appearance of uniformity, however, patent law is actually as varied as the industries it seeks to foster. Closer examination of patent law demonstrates that it is unified only in concept. In practice the rules actually applied to different industries have shown increasing divergence. As a practical matter, it appears that although patent law is technology-neutral in theory, it is technology-specific in application. The differential application of patent standards to different industries correlates with a larger theoretical confusion in patent law. While most theorists agree on the general utilitarian framework of patent law – that is, they agree on the goals the patent statute is intended to achieve – they have offered radically different ideas regarding how patent law should be interpreted to achieve those goals. We examine the various different theoretical approaches to patent law. We suggest that none of these theories is entirely correct. Neither are they entirely wrong. We show how various different theories of patent law succeed in explaining the application of patent law to particular industries, but fail when taken outside the narrow context of those industries.

The fact that economic evidence, patent doctrine, and legal theory all vary by industry leads us to question whether patent law should explicitly attempt to tailor protection to the needs of specific industries. We point out a number of risks inherent in such a technology-specific approach, particularly one administered by Congress. In particular, concerns about rent seeking and the inability of industry-specific statutes to respond to changing circumstances lead us to conclude that we should not jettison our nominally uniform patent system in favor of specific statutes that protect particular industries. Nonetheless, there are other ways the law can take account of the needs of different industries. We argue that it makes sense to take economic policy and industry-specific variation explicitly into account in applying general patent rules to specific cases. Patent law gives the courts substantial freedom to do this by means of flexible legal standards we call “policy levers.” We identify ten sets of policy levers that already exist in patent law, and the ways in which they implicitly or explicitly permit the courts to take account of different types of innovation in different industries. We also identify a variety of other places where the statute grants the courts substantial discretion, and suggest ways that those discretionary standards could serve as policy levers. Finally, having identified certain policy levers and the method of their employment, we consider the economic characteristics of innovation in five different industries that appear to be likely candidates for industry-specific rules: chemistry, pharmaceuticals, biotechnology, semiconductors, and