“No patents on seeds”, “no property rights on the natural heritage”. The opponents to patents on plant breeding methods are numerous, and they include many fringes of society that we are not used to hear in debates on technical features of patent policy. This debate is a good opportunity, and we should welcome the fact that it takes place. Often enough, important decisions in IP policy have been taken without the civil society and without triggering much attention. Nevertheless, there seems to be a lack of economic analysis and of familiarity with issues of the patent system. I try to fill this gap, highlighting four main points to analyze: Is patent protection suited for the features of the industry at stake? Can we expect sufficient patent quality? What effect would patent protection have on follow-up innovation? How would the patent system interact with sector specific regulation? In conclusion of the analysis, I come to endorse the EPO decision to refuse patentability of plant breeding methods; even though I do not recur to generic arguments such as “no profit on life”.
The EPO has decided on December 9 that biological breeding methods are not patentable, even if they use biotechnological tools (view). This is a partial victory for the movement “No Patents on Seeds”, an alliance of 300 NGOs, that has strongly opposed patent grants on breeding methods and drawn a big deal of public attention to the EPO decision earlier this month. The current decision is associated with patent EP 1069819 B1, better known as „the Broccoli patent“. This patent has been granted in 2001 to Plant Bioscience Ltd., a British Biotech Company. It protects a breeding method for broccoli in which a marker gene is used to identify the desired characteristics of the plant.
In the current discussion, an important role has been played by actors from the civil society who are usually not or not mainly concerned with Intellectual Property Rights. This is for instance the case of Greenpeace. The environmentalist NGO is cofounder of “No Patents on Seeds”, and has largely contributed to building public awareness of the issue. Thereby they confirm a recent trend that NGOs and communities draw political and public attention to issues relating to IPR. IPR practitioners, International Organizations and even politics for a long time treated these topics as purely technical, and therefore un-political. The Anti-Counterfeiting Trade Agreement, negotiated behind closed doors, is hopefully the last example of this unfortunate way of decision making. The discussion on the “Broccoli patent” has already yielded one very positive result: contrariwise to the traditional perception of patent offices as service providers for innovators, the scope of the debate and the variety of interest groups involved made it very clear that patent policy is a matter of general interest.
It is therefore very positive that there is an explicit recognition of the political nature of the issue. Nevertheless, we cannot abstract from the technical dimension of the issue when we want to come through with a convincing argument. This clearly appears from a rapid review of many of the arguments raised in the media. Many arguments of the opponents of “Broccoli patents” follow a very general stance: “No IPR on life”. This position is definitely too broad, as it is on the one hand counterfactual (there are many patents on gene engineering, and there have been Plant Breeders Rights (PBR) and other specific property rights on virtually all edible crops and plants for quite a long time), and on the other hand it lacks a clear rational justification. Other arguments of the opponents are concerned with the independence of small independent farmers and breeders, who are perceived to be threatened by mighty biotechnology corporations. It is hard to see how patents yet to be granted could prevent farmers from continuing to do what they do. After all, it is unlikely that the EPO will grant patents violating so blatantly the basic criterion of patentability, which is novelty. Finally, many opponents to “Broccoli patents” seem to be motivated by fighting genetic engineering overall. Also this is not the point here. First, each issue should be discussed on the merits of the issue itself. We should therefore not discuss plant breeding patents on the ground of our beliefs on genetic engineering. Second, patents on genetic engineering act as an upward price regulation on genetic engineering tools in breeding, they should therefore reduce the use of these tools in breeding at least in the short run. This very brief tour de force of some of the main arguments in the debate on plant breeding methods shows the need to discuss the more technical aspects of the issues at stake. There are of course technical aspects to plant breeding for which I am absolutely not an expert. But there also technical aspects in this debate that are recurrent in IP issues, and I feel that I can make a contribution to these aspects.
From the viewpoint of the economic analysis of IP, the debate around breeding method patents is just another example of extending the scope of patentability. Prototypes have been software, business methods, and the Human Genome. In all these debates, the main point to analyze is whether the particular industry to which patentability is to be extended is actually suited for this type of protection mechanism. Critics of the patent system argue that current IP policy imposes a “ One size fits all” approach, imposing patent protection on industries that used to work well with other mechanisms, such as copyright or specific protection rights.
In our case, Plant Breeders Rights (PBR) traditionally confer such a sector-specific intellectual property to plant breeders. “One size fits all” is not necessarily wrong in theory, but in practice it is likely that there are actually markets for which patents are socially and economically better suited than for others. Therefore patents should be granted only in technologies that are suited for patents, and specific property rights have their merits in other areas (copyright, database protection, PBR). Traditional plant breeding exhibits economic features that plainly justify such a specific approach: characterized by low upfront investment, extremely high cumulativeness, and a high mistake rate (innovation by try and error), the sector seems not particularly well suited to patents. PBR grant fair use of the breed and exclude reach-through claims. They therefore adapt to the specific features of the traditional plant breeding. On the other hand, genetic engineering is a typical example of patent-intensive innovation: there is a costly and risky upfront investment, which is directed towards a precise objective, and the resultant invention is easy to copy. The Broccoli Patent is therefore on an edge between two very different innovation environments. If patents were to be allowed for this intermediary sector, we would neither be in a clear-cut patent intensive industry such as pharmaceuticals, nor in a typical no-patent environment such as basic science or most types of software. In such an intermediary case, the welfare effect of introducing patents crucially depends on patent quality.
History reveals that patent quality is often dubious in areas to which patentability has recently been extended (see for instance patents in software or business methods). This is probably due to a less codified record of prior art and to missing expertise of patent officers. It is therefore likely that when extending patentability even further, a lot of patents of questionable quality would be granted. This runs opposite the declared objectives of the EPO and its “raising the bar” policy. Even if granting of low quality patents would only be a transitory problem, patents once granted can be kept in force for twenty years! Thus mining a nascent technology with low quality patents can turn to be a heavy burden.
Low quality patents are particularly costly and nasty where innovation is strongly cumulative (see the growing literature on anticommons and the patent thickets). Patents on upfront inventions do not only reduce implementation, but also the use of the technology for further research. This distinction is crucial. A patent that reduces or slows down implementation has a negative effect on welfare for the time it is kept in force. A patent that reduces or skews follow up research would have a negative effect on growth and a permanent negative effect on social welfare, even when it already has expired. This is a strong concern especially for biotechnology, which is strongly cumulative. Owners of biotechnology patents often collect royalties from follow-up inventors through so-called reach through rights, which are claims on shares of royalties for subsequent inventions drawing from the first invention. Even “the Broccoli patent” itself claims property rights on breeds made by subsequent breeders.
Finally, when analyzing the welfare effects of a policy change, one needs to take into account not only the social institution that is to be changed, but also the interactions between this institution and other features of the industry of concern. In the case of patents, it has often been analyzed that the welfare effects of patents crucially depend upon the institutional setting of industry-specific mechanisms. For instance, I and some other authors of this blog have studied into detail the interaction between the patent system and standardization in ICT industries. This interaction is highly problematic and requires intensive specific regulation. Another classical example is pharmaceuticals: the interaction between the patent system and drug administration requires strong sector specific regulation, including specific provisions of patent law. We would have to expect such problematic interactions also for plant breeding. There is detailed public regulation of the plant breeds that are allowed for human alimentation, and there is strong regulation of the allowed scope of gene engineering. Both these regulations are very vulnerable to capture by special interest groups. Regulation is said vulnerable to capture when a small number of players concentrates important material interests in a specific regulation, and when there are strong asymmetries of technical expertise. Both these features are given for plant breeding and genetic engineering, and it is likely that stronger property rights can only exacerbate the capture of public regulation.
An economic analysis of the industry thus leads me to endorse the EPO decision to refuse patentability of plant breeding methods. This analysis does not require arguments or motivations that are outside the patent system. Rather I have tried to show that the point of the opponents of “the Broccoli patent” can be made on the merits of the case, and by reference to the rules and reasons of the patent system itself.