Totally Evil Patent
A hypothetical Patent whose purpose is to demonstrate that the patent system is morally unsound.
The Totally Evil Patent has three important properties:
- Validity Property: The patent is valid according to the criteria of novelty, non-obviousness, usefulness and serving a moral purpose.
- Low Value Property: The value to society of the disclosure of the invention described in the patent is low. The extreme value of this property is the Zero Value Property, i.e. the value to society of the disclosure is zero.
- High Cost Property: The cost to society of granting the patent is high. The extreme value of this property is the Infinite Cost Property: an example would be when the granting of the patent gives the patent holder the right to cause the cessation of all human life (for example because it turns out that everyone needs to use the patented invention to carry on living).
Actual patents may have some or all of these properties. The Validity Property is important for this discussion: if a patent is not valid then it can (in principle) be overturned within the existing system. Thus the existence of invalid patents with low value and high cost does not indicate any need for patent reform. In general a patent is considered "evil" if its cost greatly exceeds its value.
A Valid Patent with Zero Value
The Validity Property implies that an invention is non-obvious at the time the patent application is granted. Thus there is some non-zero value to society if this invention is disclosed at that time. However the value of disclosure at the time of granting can be arbitrarily low. A practical issue is that the patent system does not require a patented invention to be disclosed until the patent is granted. A patent has Provable Zero Value if someone else discloses the same invention after the application is made but before the patent is granted or otherwise publicly disclosed.
A patent may still have zero value in a case where the invention is disclosed by another party after the patent is granted. In this case the zero value property is unprovable, because one can always claim that the second disclosure was made by someone who had read the description of the invention from the patent. Being unprovable is not the same as being provably false: in particular we have to consider the possibility that someone else would have eventually invented and disclosed the same invention, even if the patent applicant had never invented their invention at all.
At any given time, the value of a patented invention and the cost of its suppression both depend on prevailing circumstances, and circumstances can change. In the modern world circumstances are changing much more rapidly than they did at the time when patents were first invented (12th century Venice?). In particular, circumstances affecting the value and cost of a patent are very likely to change within the period that a patent is granted for (currently 20 years from application date in most countries). Of course one of the major causes of these changing circumstances is the invention of new technologies.
A Valid Patent with Infinite Cost
Ignoring the possibility of prolonged torture or unbearable agony (arising from whatever cause), the highest cost that an individual can pay for anything is their own life, and – if we take a biological view – the life of all of their descendants and close relatives. A patent holder could exact this cost from an individual if the individual required use of the patented invention in order to save their life and the lives of all their living relatives, and if the patent holder refuse to license the patent to that individual, or if they set a license fee for use of the patent which the individual could not afford to pay.
Interestingly enough, patents for AIDS drugs are a real-life example of patents that effectively have infinite cost to large numbers of people who cannot afford patent license fees. A more recent example is the set of patents that would be required to be licensed to enable measures that would prevent the catastrophic spread of a mutated version of the bird flu currently spreading in Asian countries (as at Jan 30 2004). It is assumed that the patent system would be ignored when it became obvious that millions of lives where at stake, but the problem is that by the time we realise this it may be too late to prevent a disaster.
These medical examples do not satisfy the Zero Value Property; indeed a non-trivial amount of effort and investment was required in order to invent the technologies covered by the patents involved. Nor do they have the Infinite Cost Property with respect to the human race as a whole – the sheer size and genetic variation of the human population makes it very unlikely that any disease could exterminate the human species. For example, 1% of the population is highly resistant to infection by the AIDS virus in its current form. The worst plague in known history, the Black Death, did not kill more than 50% of the world's population.
Software and "Problem" Patents
The real life patents that come closest to the Zero Value Property are mostly software patents, and in particular so-called "Problem Patents". A Problem Patent is a patent that discloses an obvious solution to a non-obvious problem. Although potentially invalid due to the obviousness of the solution, the non-obviousness of the problem is usually enough to protect the patent applicant from the main proof of obviousness: prior art. The evilness of such patents is most obvious in the case of software patents, as the cost of writing and distributing software that implements an "invention" is often close to zero (at least calculated on a per user basis), especially given the use of the Internet and World Wide Web to enable the cooperative development of open-source software.
Implications for Patent Reform: Ad Hoc Revocation
Proving the possible existence of a Totally Evil Patent does not prove that all, or most, or even a substantial portion of patents actually granted are evil. It may be that almost all patents have a net benefit to society that is either zero or positive. What the concept of the Totally Evil Patent does prove is that it is possible for just one patent – or a very small number of patents – to have a net cost to society that is so large that it cancels out all the positive benefits of the other patents.
This leads to a simple suggestion for reform of the patent system: allow for the ad hoc revocation or compulsory licensing of a small portion of patents that are granted. The process for selecting patents to be revoked has to be purely political – either a choice by referendum, where everyone votes for which patents to revoke, or if that is too hard, a "Ministry of Ad-Hoc Patent Revocation". The ministry would choose which patents to revoke, explain the choice, and if the public didn't like it then they would have to vote in a new government which would then appoint a new Minister of Ad-Hoc Patent Revocation.
The reason that patent revocation has to be a political process is that there is no completely objective procedure for determining that a patent is evil. It is hard enough trying to prove the obviousness of an invention at the time of application; it is quite impossible to prove that an invention would or would not have become obvious at some later date if it had not been disclosed when it was disclosed. It is also impossible to measure how much opportunity has been lost by giving the first inventor of an invention the right to suppress the use of that invention for twenty years. All of these things can only be guessed.
Necessarily the patent revocation would have to be done on a per-country basis. Thus your patent may be unlucky enough to get revoked in the US, but perhaps the inhabitants of France are OK with your patent, or regard other patents as having a higher priority. Revocation may not extend to a full patent, for example decisions could be made to revoke just those claims considered to be too broad.
Pros and Cons of a Revocation System
There are those who believe that the patent system is sacred, and that it must not be diminished in any way. They will argue that a revocation system will fatally undermine the protections that the patent system gives to inventors. However, if a quota is set to a specific percentage of patents granted (e.g. on a yearly basis), then the risk to the individual inventor is quite small. For example the percentage might be set at 0.01% – in a country that grants 1,000,000 patents a year, that is the revocation of 100 patents a year. To accurately calculate the effect this has on a random patent holder's income, we must compare this percentage to the probability that a patent generates significant income for its holder. To continue the example, if only 1% of patents ever generate income, then revoking 0.01% of evil patents is likely to result in a revocation of 1% of income-generating patents (most evil patents are probably also income-generating patents).
Revocation does not have to imply cessation of licensing income, as revoked patents may be compulsorily licensed, or a one-time buyout fee may be paid. However such a compensation system is open to abuse if patent-holders can be given income in return for patent revocation, and there is no way to prove how much licence fees they could have collected anyway. It may be a good idea for each country to set a total budget for patent revocation, and to award buyout payments in proportion to the number of votes made to revoke each patent.
The revocation percentage can also be adjusted as necessary. If it turns out that revocation is successful, in particular that countries that revoke certain patents benefit in an obvious way from those revocations, then that will be good reason for those countries to consider increasing the revocation quota to some larger percentage. The quota may or may not reach 100% in the long term. The important thing is that a sliding revocation quota represents a conservative and gradual approach to winding back the patent system: it avoids the FUD that is raised (by defenders of the patent system) against proposals to completely disallow certain types of patent (such as software patents).
Politically driven patent revocation would also appear to undermine the "rule of law". However patents themselves are individual "laws", and patent law is in some sense a "meta-law". It can be argued that democracy requires citizens to have the right to vote for those who make their laws, or even to vote on the laws themselves, and that this right should therefore extend – at least to some degree – to patents.