Patents: The worst incentive scheme for innovation, except for all the others?

A demonstration took place in Berlin in front of the Chancellor’s Office on June 13, 2021, with the slogan #GebtDiePatenteFrei, calling for patent protection on vaccinations and medication to be abolished.

The initiative was supported by several alliances, such as ATTAC Germany, medico international, and ZeroCovid.org. There was also a public petition running with the title “Kill Patents – revoke patent protection on all essential medicines.” While political circles are currently discussing lifting specific patents on COVID-19 vaccines, these initiatives seek to have patents on entire product groups lifted, or indeed to abolish the very patenting system itself.

Calls for abolishing the patenting system are not new. As the patenting system has grown increasingly significant in the wake of 19th-century industrialization, dissatisfaction and public debate have also increased and major demands arise to abolish patenting (Khan 2013, Machlup & Penrose 1950).

Opponents of having a patent system usually refer to the welfare losses resulting from the market power of patent holders. They also allege that the system is abused by dubious patent trolls who don’t produce any products but create a threatening backdrop for manufacturers to collect seemingly unjustified license fees. (It often goes unmentioned that this phenomenon might just result from a division of labor, in which an inventor prefers to sell licenses rather than investing in factories.)

Surprisingly, debates on the advantages and disadvantages of the patent system raise rather few new ideas. Patents can lead to temporary market power for the patent holder. Non-producing patent entities, including patent trolls, are also not a new phenomenon, as Lamoreaux and Sokoloff (2001) show using the example of the 19th century. As early as 1950, when the patent system was (once again) hotly debated in the US, the economists Fritz Machlup and Edith Penrose argued that “despite all the changes in the economic scene, our thinking on the subject [the usefulness of a patent system, author’s note] has hardly changed over the century.”

Any viable patenting system has its disadvantages, but there is no better alternative for promoting private innovation work. Seen in this light, the patent system is (paraphrasing Winston Churchill) the worst system for encouraging innovation, except for all the others that are tried from time to time.

Why do we need innovation incentives at all?

Innovations usually have some social benefit and are often built around technical inventions. Knowledge and inventions have specific economic characteristics; in contrast to physical goods, they have the features of a public good. First, an invention, or the idea behind it, can be replicated virtually without cost and thus can be made available to an unlimited number of users (non-rivalry in use). Second, it is impossible to exclude third parties from using an invention as soon as the idea behind it is publicly known (non-excludability of third parties).

These specific issues result in an incentive problem, which, if it isn’t resolved, will lead to an “underproduction” of technical innovations compared to a social optimum (Nelson 1959, Arrow 1962). While the replication of an invention is virtually free, the research and development are risky and costly. If competitors cannot be stopped from using a functioning invention, there is little incentive for private actors to take this risk and invest in knowledge production. Cheap imitations by competitors would stop inventions from being put on the market with a substantial profit margin. An underproduction of inventions thus occurs, and this slower technological progress leads to a loss in prosperity.

Patents as an incentive for innovation: A double-edged sword

Patents can solve the incentive problem outlined above, because a patent holder can exclude third parties from using the patented invention for a limited period (usually 20 years). In return, the patent holder must publish details of how the invention works so that third parties can replicate the invention after the patent protection expires. This exclusion results in market power (and, in extreme cases, a de facto monopoly) for the duration of the patent and allows for pricing above the manufacturing cost (the price level in perfect competition) for profit. The prospect of this profit creates incentives for private investment in risky and expensive innovation projects, from which the public benefits (Arrow 1962).

The German biotechnology company BioNTech, for example, has invested for over ten years in developing drugs based on mRNA molecules. They can make a profit after the approval of their COVID-19 vaccine because the active ingredient is under patent protection. The development and subsequent availability of the vaccine benefits the public.

This example also shows the main disadvantage of patents: increased prices resulting from the market power of the patent holder restrict access to patented products for some – especially low-income – groups of buyers. While this is bearable for some products, restricted access to essential medicines can have life-threatening consequences for patients. This argument has been put forward by opponents of the patent system continually over the centuries (Khan 2013, Machlup & Penrose 1950).

Empirical evidence for the incentivizing effect

The welfare-enhancing incentive for private innovation and the temporarily limited welfare loss counterbalance each other. The essential question is therefore which consequences outweigh the others: welfare increase resulting from a rise in available inventions or welfare loss due to exclusivity rights?

Welfare losses are often clearly identifiable. For example, in the COVAX initiative, the WHO emphasizes that COVID-19 vaccines are hardly affordable at their current prices in countries with a low economic output. However, the connection between the level of innovations produced and the existence of a patent system is difficult to identify empirically.

The studies that are available show that just a weak correlation can be empirically identified between patents and resulting innovation incentives. Across various industries, the strength of patent holders’ rights correlates with the number of registered patents and R&D expenditure (Ohlhausen 2016). The pharmaceutical industry shows that longer patent terms are associated with increased innovation efforts (Budish et al. 2015, Gaessler and Wagner, forthcoming). While patents in the chemical and pharmaceutical industries are the basis for the commercially successful distribution of inventions and the primary incentive for research activities, surveys have shown that patents play a less important role in other industries. There, what counts most are confidentiality, lead time, and complementary assets, such as manufacturing capabilities or customer loyalty (Levin et al. 1987, Cohen et al. 2000).

Are there any alternatives?

Empirical evidence for the positive incentive effect of patents is extremely difficult to get. So what alternatives are there? And would they be more desirable than a patent system?

An obvious approach is to invest directly in research work, as with tax-funded universities and public research institutes. It has been shown that – despite their valuable research – these institutions cannot commercialize the resulting technologies on their own. Private actors are needed, and their support requires incentives.

Subsidies and tax benefits for research-based companies are also financed by the public. Research shows that there is indeed an innovation-promoting effect when subsidies are added to an existing patent system. It is unclear to what extent they could cushion the abolition of the patent system and how far they would have to be extended for this purpose. The resulting costs would have to be distributed across all taxpayers, regardless of whether they would benefit.

Prize money for solving pre-specified problems is also a research incentive. Because the inventor relinquishes the rights to the invention, there is no welfare loss due to price premiums. A closer look at prizes reveals several drawbacks (Galasso et al. 2018, Wright 1983). Prizes assume the problems are known. This is not always the case, and many problems do not have prizes offered for them. Moreover, the value of solving a problem must be determined beforehand to assign prize money accordingly (which must also be pre-funded). In practice, this is difficult. The award does not always follow strict performance criteria but is evidently characterized by nepotism, and the quality of the solutions that are submitted is often poor (Khan 2011). Patents do not have these issues – problems do not have to be named, and market success determines their value.