A Replacement for the Patent System

Patents as originally setup were a deal between a government and someone who had a good idea. The person with a good idea would get a patent providing him with the right to make a profit on all use of the idea for some period of time and, in return, make the idea available for others to use. Very significant changes have occurred since the inception of the idea with all of these changes making the basic deal favor the patenter over society.

The pace of innovation has dramatically increased to a time scale as short as 1 year. Patents, with a fixed time scale of 17 years are now a radically better deal for the patenter because the next 10 or so generations of technology which build upon a patented idea must pay homage to the original patent owner. Consequently, new technology now tends to be heavily encumbered with multiple patents to which licensing fees must be paid creating an administrative nightmare as well as possible monopolistic choke points on each patent.

The standards of what is a patentable idea have also steadily relaxed. Software patents were originally rejected but now are commonplace. Business process patents are another new variety which have significant implications. The issue of "stupid patents" has become especially bad in both of these new areas where many patents are made based upon the kind of original thinking that would be expected of a competent software designer who thought about the problem for a few minutes. Examples include a patent on "windowing" to avoid the y2k bug and a patent for the process of sending audio data over the internet. Many others exist and have been detailed elsewhere.

The latest form of patent is on a business process itself. If you regard a company as a machine, the patent office now excepts patents on pieces of this machine. It is difficult to imagine such patents having any value for society at large and easy to imagine destructive potential in these patents. Business process patents are often not licensed for use by other companies, but are instead kept by the company for its own (exclusive) use. The consequence is that every other company is forced to perform in inefficient ways which do not copy the patent. One example of this is the Amazon patent on "One Click Shopping". The basic idea is to make it very easy for customers to buy items over the web and the means is trivial. How does it give society any benefit to give an exclusive monopoly on "One Click Shopping" for 17 years? It doesn't.

We are stuck in a conundrum. The idea of compensating people for good ideas is a good idea because people tend to be very self-motivated in general and it is best for society to harness this motivation for the improvement of society. However, the current system is inflexible, poorly managed, inefficient and appears to have entirely lost sight of the fundamental deal which society made in creation of the patent system.

Inflexibilty: 17 year patents at a minimum granting essentially all right to the patent holder. The inflexibility could easily be improved by passing a law making the terms of a patent more negotiable.

Poorly managed: As evidenced by the preponderance of stupid patents. With the cost of a patent application in the $10000 range it is pretty clear that a nontrivial amount of income for the patent office is generated in patent application fees. It is unsurprising to learn that the patent office has been succesfully encouraging patent applications by gradual degradation of patent standards and widening of scope. The fix for this problem is more difficult. Fundamentally, patent granters need to be motivated on the behalf of society in order to conduct a meaningful negotation over granted patent rights. While general improvement of society may be of some concern to the negotiators, a more direct system of reinforcement is likely to be useful. Lacking any better idea, I would propose a patent tracking office which tracks and attempts to predict the impact on society of granted patents over time. The original patent negotiators can then recieve bonuses based upon the "performance" of their patent negotiations. While this is a difficult task, it is also rewarding in its own right because better knowledge of the impact of patents on society would aid in conducting patent negotiations.

Inefficient: Legal enforcement of patents and negotiation are a very significant burden on patent holders. This problem is fundamental to the way patents work. There is a very unsettling nonlocality to the patent system. If Jim gets a patent then Jill can not legally use the idea without paying Jim royalties. Even if Jill never heard of Jim's patent and developed the idea on her own this is true. Indeed, the number of patents is now so large that it is now difficult to know if you are violating a patent. Further complicating the matter are court cases which suggest that only a patent lawyer is qualified to determine whether a patent is in violation. It is remarkably inefficient for all parties to bear the burden of this nonlocality and legal maneuverings. Correcting this problem is not simple because it is built into the way patents are done. If instead of granting exclusive rights the government does upfront payments on the behalf of society to ideamakers these inefficiency issues may be avoided.

So, in total the proposal is to increase the negotiation flexibility available to the patent office, create a seperate patent impact oversight office with which to determine the impact on society of patent grants, use the patent impact studies to provide bonuses to good patent negotiators and to replace the standard monopoly style patent with direct compensation.


jcl@cmu.edu