Until recently, patents were not available for certain technologies such as computer software or methods of doing business. Today, in the U.S., patents are being granted for existing products or processes implemented via new technologies such as the Internet. Are many of these inventions obvious and therefore not suitable for patent protection?

In this article, Glynn Lunney reviews the history of the obviousness standard and its transformation under the Federal Circuit (the appeals court for patent cases established in 1982). He argues the standard was significantly relaxed, and he attributes this to a change in the Court's underlying theory of patents. He proposes to recast the obviousness standard in a way that reconciles the Federal Circuit's theory with the more traditional theory of patents.

Describing the Problem
For 150 years, U.S. courts have ruled that inventions that are trivial advances over existing knowledge do not qualify for a patent. The 1952 Patent Act precluded the granting of patents for inventions that "would have been obvious at the time the invention was made to a person having ordinary skill in the art." In the 1966 decision, Graham v. John Deere, the Supreme Court articulated the judicial test for nonobviousness, adding that certain considerations (commercial success, failure of others, or long-felt but unsolved need) may also be relevant. After 1982, the Federal Circuit elevated the importance of these secondary considerations in its decisions.

How important were these changes? A chart based on Lunney's data (see figure**), shows that by the 1990s a much smaller share of patents were found invalid on obviousness grounds. Lunney emphasizes that care should be exercised in interpreting such data, but the trend is consistent with a substantial change.

Competing Theories?
Lunney argues the Federal Circuit adopted these changes because it subscribes to a different theory of the purpose of patent law. Historically, courts have interpreted patents as a bargain that provides inventors with incentives to discover and disclose their inventions, paid for at the cost of a temporary monopoly in the market place (he calls this the bargain theory). Under that theory, the non-obviousness standard ensures that society gets sufficient benefits to justify the deadweight losses associated with monopoly pricing.

Lunney contends the Federal Circuit now prefers what he calls the simple property theory of patents. Under that theory, patents are no different than any other form of property. What matters is that patents help allocate scarce resources--the time and effort of scientists, engineers, etc. Under this theory, the patent office is equivalent to a land title agency and there is no compelling reason to consider a stringent non-obviousness requirement.

A New Requirement of Nonobviousness
Lunney reconciles these theories with a normative economic argument. Patents are valuable when they encourage socially valuable inventions that would not otherwise occur because they can be easily imitated. Conversely, patents should not be granted for inventions that are not socially valuable or that would occur anyway (because they cannot be easily imitated). This approach minimizes losses associated with monopoly pricing (bargain theory) and it efficiently allocates scarce creative resources (simple property theory).

Lunney proposes a new rule: if in the absence of a patent, the inventor cannot recoup his or her investments in creativity, then the invention is nonobvious and therefore patentable (assuming the other requirements of patent law are satisfied). In practice this cannot be observed with any precision before the fact. So Lunney suggests a proxy: the share of creative activities in the total cost of bringing a product to market. To determine obviousness, that share would be compared to some standard, but Lunney concedes the standard cannot be reliably specified.

Lunney also recognizes there is an issue with the relative productivity of inventors. A highly skilled inventor will have lower invention costs and would therefore have a harder time getting a patent.

This approach also relies on some strong assumptions. Lunney assumes that private benefits are always less than social benefits, but this is not necessarily the case. Also, Lunney assumes that such a test can be applied more easily than the admittedly vague standard articulated in Graham v. John Deere. But it's not clear this is true.

In the end, Lunney articulates a new standard that appeals to our normative understanding of the role of patents.

*The views expressed here are those of the author and do not necessarily represent the views of the Federal Reserve Bank of Philadelphia or the Federal Reserve System.

**This figure represents the data from Lunney's Figure 1 (Percentage of Patents Held Invalid Where Validity at Issue and Decided) multiplied by his Figure 2 (Percentage of Invalid Patents Found Invalid For Obviousness). Thanks to Brian Kahin for this calculation.