Toward an Optimal Patent Regime for Canada

Introduction

he traditional view on patents is that they trade off static welfare losses for dynamic welfare gains (Arrow 1962; Nordhaus 1969). On the one hand, the higher monopoly prices generate deadweight losses. On the other hand, patents incentivize innovation, which can drive welfare gains. Under this view, establishing the optimal patent policy is tantamount to determining the ideal trade-off between these. Yet, the premise that patents promote innovation is increasingly being challenged (Boldrin and Levine 2013). In particular, the possibility that patents do not promote innovation — and even that they may stifle it — arises when one accounts for the cumulative nature of innovation and the fact that national patent regimes do not exist in isolation, but rather are embedded in the global intellectual property (IP) framework. This essay discusses these topics and their implications for Canada.

Cumulative Innovation

Innovations do not happen in isolation. Rather, they typically build on previous innovations. Cars were not possible before the invention of the combustion engine and computers required the Shockley transistor. It is, therefore, important to account for the cumulative nature of innovation in examining the effects of patents on innovation.  

When one patent builds on the innovation of another, both the initial and follow-on innovators have blocking rights on the use of the new innovation. That is, the new innovation cannot be commercialized without the consent of both parties. To the extent that the follow-on innovator can bargain ex ante (before investing to develop the idea) to acquire a license to use the initial innovation, the follow-on innovation will be pursued whenever it is value generating (that is, when the profit that it generates is larger than its cost to develop). However, when bargaining can only occur after the development of the follow-on innovation, the follow-on innovator will sometimes not pursue profitable innovations because they will be held up for their investment. Consider, for example, the case where the follow-on innovator has to invest $3 million to develop an innovation that generates $4 million in profits. Prima facie, the inventor would be foolish not to develop since the innovation generates a net profit of $1 million. However, the inventor knows that following development, the initial innovator would demand half of the $4 million since both innovators hold equal blocking rights on the use of the follow-on innovation. Fundamentally, the net $1 million value of the innovation will be split as $2 million to the initial innovator and –$1 million to the follow-on innovator, so the latter would choose not to pursue it. From society’s point of view, a welfare-increasing innovation was not pursued because of patents and the inability to bargain ex ante.

When are we likely to see less follow-on innovation due to an inability to bargain ex ante? First, this is likely to occur in situations where the follow-on innovator is not aware of previous patents that they may be infringing on. Engineers have incentives not to conduct a thorough review of the prior art so as avoid the larger potential damages awarded for willful infringement. And, even if they do conduct such a review, in the case of complex innovations with multiple technological inputs, they know they are likely to miss at least some of the previous patents on which they may be infringing. To further complicate matters, patents are often open to interpretation as to what they cover and even whether they are valid. As reported by Mark Lemley and Carl Shapiro (2005), nearly half of fully litigated patents are declared invalid. Given the real possibility of being held up for their investment after they develop their idea, inventors may rationally choose to forego innovation altogether.

Second, ex ante bargaining can also break down if the two parties do not agree on the value of the follow-on innovation and they therefore cannot agree on a fair split of the surplus. A third situation where it may, in practice, be impossible to bargain ex ante is when the technology is so complex that there are too many patents on which the new innovation would infringe, and too many parties to negotiate with. Trying to work out licensing agreements ex ante with each of the parties could delay development to the point where, once developed, the innovation would be obsolete. Alternatively, if the follow-on innovator cannot coordinate bargaining among all the parties, it could lead to a stacking of the royalties, potentially reducing their profit to the point of blocking development altogether (Heller and Eisenberg 1998). 

One way that large firms deal with such patent thickets is by setting up patent pools, where firms cross-license all the patents to other members of the pool. Of course, smaller players are frozen out since they do not have the patents to contribute to such pools. Thus, patent rights could severely impact follow-on innovation by all parties, but the effect could be especially felt by smaller firms.  

Overall, different theoretical models on cumulative innovation provide different predictions on whether patents promote or hinder cumulative innovation. It is, therefore, perhaps most productive to turn to the emerging empirical literature on the topic. Here the evidence, while still inconclusive, suggests that patents stifle follow-on innovation. Fiona Murray and Scott Stern (2007) find that when the science in a journal article gets patented, those articles see a 10–20 percent decline in citation rates. Heidi Williams (2013) finds that genes that are subject to IP spur 20–30 percent less follow-on research and development. Alberto Galasso and Mark Schankerman (2015) find that when a patent is invalidated by the courts, it leads to a 50 percent increase in citation rates, in particular, in the sectors of computers, electronics and medical instruments. They further find that the effect is driven primarily by the invalidation of large-firm patents and the subsequent increase in citations to those patents by small firms. In contrast to these results, a recent paper finds that the patenting of human genes has no effect on follow-on scientific research and commercial investments (Sampat and Williams 2015).

Overall, while the existing evidence remains inconclusive, it seems entirely plausible that patents could stifle innovation when the cumulative nature of innovation is taken into account. This is particularly true for innovation by smaller firms and in complex sectors of the economy where products combine many different inputs. 

Patents in a Globalized World

To better understand the impact of a national patent regime on domestic innovation, it is also important to consider the international context. The 1883 Paris Convention for the Protection of Industrial Property enshrined the principle that inventors should receive “national treatment” for intellectual property rights (IPRs) in all signatory countries. This same principle was subsequently embedded in the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), ensuring that inventors can protect their IP in all major markets around the world. Inventors therefore derive increased incentives to innovate not just from the prospect of a patent monopoly in their home country, but also from the prospect of receiving a patent in foreign markets (Park 2012; Blit and Zelaya 2015). When deciding whether to develop an idea into an innovation, private agents therefore weigh the cost of development against the total profit that they would derive across all countries from the innovation. 

In the case of a smaller economy, then, it is not clear that its national patent regime will have much impact on the incentives of domestic (or foreign) inventors.¹ In fact, under national treatment, all countries have strong incentives to free ride on the patent protection of other countries, since these can give inventors sufficient incentives to innovate while not generating deadweight losses domestically. Partly in response to this, numerous international agreements have been put in place to ensure minimum harmonized standards for patent protection, with TRIPS being the most notable. But such agreements are controversial. In particular, countries will not, in general, agree on the optimal harmonized level of patent protection, with more innovative countries preferring stronger protection than less innovative countries. In fact, more innovative countries would choose a level of protection that would exceed that which maximizes joint welfare. It is not surprising, then, that the TRIPS agreement, which was championed by the more innovative countries, is seen by many as benefiting the more innovative countries at the expense of the poorer, less innovative ones. In fact, it may even be that, as a result of extensive lobbying by the US copyright and patent lobby, the level of patent protection enshrined in the TRIPS agreement is beyond what is optimal, even for the highly innovative United States. 

We should, therefore, be gravely concerned that IPRs continue to be ratcheted up through bilateral trade agreements that contain “TRIPS-plus” standards, called this because they establish higher standards than those mandated by TRIPS. For example, such agreements have extended patent terms to longer than 20 years,² implemented more stringent restrictions on the compulsory licensing of patents, extended IP protection to more subject matter and mandated stronger enforcement mechanisms.  

Moreover, because the TRIPS agreement has a most-favoured nation (MFN) clause,³ presumably any concession made to another country within the context of a bilateral trade agreement (such as stronger restrictions on compulsory licensing) applies to all other World Trade Organization (WTO) member countries. The MFN clause makes sense within the context of removing trade barriers, because economic theory says that overall welfare is highest with zero barriers. However, the same is not true for patent protection, where there is an optimal level and any increases beyond that are welfare destroying. In spite of an emerging consensus that patents are likely already too strong, they continue to be increased through bilateral trade agreements and multilateral agreements such as the now-defunct TPP agreement.

The Canadian Context

Many studies have looked at the impact of patent protection on innovation for the United States (Jaffe 1999) and Japan (Sakakibara and Branstetter 1999), and across a panel of countries (Kanwar and Evenson 2003; Allred and Park 2007; Qian 2007; Kyle and McGahan 2012; Blit and Zelaya 2015). While some of these studies find a positive relationship between patents and innovation, others find no relationship. To the best of the author’s knowledge, no studies have effectively addressed this question for Canada. This is a clear gap that must be addressed if Canada is to develop an effective national patent rights policy.  

Fortunately, some inferences can be drawn from the theory. While the simplest models suggest a causal relationship between patents and innovation, the prediction becomes less clear when assumptions are relaxed and the cumulative nature of innovation is explicitly modelled. To minimize the risk of patents stifling follow-on innovation, patents should be as transparent as possible so follow-on innovators can acquire any licenses before they invest in their innovation.  Patents should be both easy to search and understand. They should also be narrower and more clearly demarcated so future innovators have a clearer understanding of what prior patents they are likely to infringe on. The minimum usefulness, novelty and non-obviousness criteria for obtaining a patent should also be strengthened to minimize the number of marginal patents that themselves contributed little real innovation, but can clog the system and inhibit future innovation.

A further prescription arising from the theory is that as a relatively small economy, there is little benefit for Canada to implement strong patent rights, beyond the requirements to meet its international obligations. The theory suggests that a strong Canadian patent regime would generate significant domestic deadweight losses and higher prices for all Canadians, while contributing little to promote Canadian innovation. What little empirical evidence there is for Canada seems to support this conclusion. As of 2014, fewer than 12 percent of the patents granted by the Canadian Intellectual Property Office (CIPO) were to Canadian inventors. And, Canadian inventors are unlikely to be swayed by the strength of domestic patent protection. Indeed, statistics presented by Joël Blit (2017) suggest that for Canadian inventors, patenting domestically is likely an afterthought. Canadian inventors take out between three and four times more US patents than Canadian patents, and the large majority of innovations that they patent in Canada are also patented in the United States. The latter finding, however, is based on the examination of a small sample of patents. Further research is warranted to better understand what role, if any, CIPO plays in promoting domestic innovation and, hence, whether a reform might be beneficial.

The patent regime must be a central consideration of any modern innovation strategy because patent rights influence knowledge diffusion, access to innovations and firms’ incentives to innovate. In the case of Canada, it does not appear that stronger patent protection would foster increased innovation and, at least for follow-on innovations, there is every possibility that they would hinder them. As such, Canada should be cautious about signing trade agreements that further ratchet up patent protection. The adoption of ever-stronger patent regimes runs counter to Canada’s historical emphasis on balancing the interests of inventors and users.