"...like everything else the legal system touches nowadays, U.S. patent law has been hijacked so that it now operates nearly in reverse, deterring research and penalizing innovation...

"RIM faces a possible injunction barring it from providing BlackBerry service in the U.S., as well as $450 million or more in payments to the plaintiff, NTP, whose only significant assets are the patents it claims RIM has infringed.

"In the RIM case, NTP offers no product that competes with BlackBerrys. It sells nothing at all."
-- an editorial in the Wall Street Journal

Intellectual property law is increasingly important. In this essay, I am going to propose that the patent law be changed to incorporate something analogous to the "fair use doctrine" in copyright. Bear in mind that I am an economist, not a lawyer, so my treatment of legal issues will have an amateur character.

My understanding of "fair use" is that it puts some flexibility into copyright law. For example, it would be illegal for me to copy an entire editorial from the Wall Street Journal, but the excerpt that I quoted above constitutes fair use. Another fair use might be copying the music I bought on a CD onto a hard disk so that I can listen to it in a different location. Unfair use would be making copies on CD's and selling them to strangers.

Some of the problems with patent laws could be fixed by developing standards for "fair use." Under a "fair use" standard, there would be circumstances where one company could use another company's patented ideas in a way that constitutes limited infringement (with a low ceiling on damages) or no infringement at all. For example, a patent adjudicator could take into account the fact that the patent holder has not made an effort within a reasonable time period to market a product incorporating the patent.

A "fair use" doctrine for patents could take into account the difficulty of developing the idea for the patent. For example, I think that most of the ideas that I have seen for stopping spam -- Bayesian filters, forcing senders to manually enter codes, schemes to charge senders, etc. -- could have been arrived at ten years ago by a group of geeks sitting in a one-hour brainstorming session. On the other hand, Visicalc, the original spreadsheet software, was a truly creative leap.

The "fair use" doctrine also should take into account the difficulty of implementing an idea. If an idea is very difficult to execute successfully, then it should be "fair use" for someone other than the originator to try to implement the idea. It is on this issue that I believe that "fair use" would support RIM, the creators of the Blackberry. Coming up with the technical concepts involved in creating a mobile communication device is only a small part of the business problem. The challenge is in making correct decisions about many subtle characteristics of implementation and design. The fact that the patent-holder has no product further demonstrates that the value of the Blackberry is in the implementation, not in the intellectual property that was patented.

A "fair use" doctrine would allow a patent adjudicator to ask the following questions:

• How much research and creativity were required to develop the idea that was patented?

If it is an idea that any number of people could have come up with in a matter of a few hours, then there should be wide latitude for fair use. A Bayesian spam filter would be an example. On the other hand, if it is an idea that has required years of testing and refinement, such as a video game platform, then the latitude for fair use would be narrower.

• Once one has the idea, how difficult is it to incorporate into a marketable product? If the product is complex and addresses a highly competitive market, as is the case with the Blackberry, the latitude for fair use ought to be wide. On the other hand, once the formula for a drug compound has been published and its efficacy and safety established, manufacturing and distribution are relatively easy. On this basis, fair use for drug patents would be rather limited.

The "fair use doctrine" would continue to protect ideas that are difficult to develop and easy to execute. However, it would provide less protection to concepts that are not embedded in working products.

The "fair use doctrine" would reduce the incentive for companies to file for patents on ideas solely for the purpose of staking claim to an idea. Instead, patents would have value only as protection for actual marketable products.

The "fair use doctrine" would protect innovative companies from having to defend themselves against patents of which they were not even aware. It would reward engineers and marketers, rather than patent attorneys.

Arnold Kling is author of Learning Economics.