What Happens When Intellectual Property Law Revises to Favor Open Knowledge

Intellectual property law rests on a specific theory of human motivation: that people create and invent when they can capture sufficient private benefit from doing so, and that exclusive rights are the mechanism through which private benefit is secured. Patents grant temporary monopoly on inventions. Copyrights grant long-term exclusive control over creative expression. Trademarks protect brand signals. Trade secrets protect undisclosed information.

The theory is not crazy. Some forms of knowledge production require large upfront investment with uncertain returns, and without some mechanism to capture returns, investors will not fund the investment. The drug development case is the canonical illustration: developing a new pharmaceutical compound, running clinical trials, navigating regulatory approval — this process costs, by common estimates, over a billion dollars per drug. Without patent protection, a competitor could simply copy the approved drug, price it at manufacturing cost, and capture the market. The original developer, unable to recoup R&D costs, would be bankrupt. The next drug would not be developed.

But this theory, taken as universal, proves too much. It implies that all the software that runs the internet — which was largely built without exclusive rights through open standards, academic publishing, and open-source contribution — should not exist. It implies that science itself, which operates on a norm of publication and open exchange, should collapse. The theory describes a real mechanism in some contexts and fails to describe reality in others.

The history of intellectual property law is a history of expansions — longer copyright terms, broader patentable subject matter, stronger international enforcement — driven by incumbent industries that benefited from the existing exclusivity and organized to extend it. The 1998 Sonny Bono Copyright Term Extension Act, which extended copyright terms by twenty years and was widely characterized as protecting Mickey Mouse for the Disney Corporation, is emblematic. The expansion of software patents into algorithms and business methods, despite significant economic evidence that this slows rather than accelerates software innovation, is another. The trajectory of IP law over the twentieth century is not a story of legislators carefully calibrating incentives to maximize knowledge production. It is a story of rent-seeking by holders of existing IP assets.

The open-source software movement is the most complete empirical demonstration that the exclusivity assumption is context-dependent rather than universal.

The Unix operating system, developed at Bell Labs in the 1970s, was originally shared openly with universities because AT&T could not legally sell it as a product. Richard Stallman's GNU Project, launched in 1983, began building a free Unix-compatible operating system on the explicit philosophical premise that software should be freely shareable. Linus Torvalds's Linux kernel, launched in 1991, used Stallman's GNU General Public License — the "copyleft" instrument that used copyright law to require that derivative works remain open. These were not primarily economic experiments; they were politically motivated challenges to the proprietary software model.

The economic outcome was unintended. Open-source software turned out to be extraordinarily productive. The reasons are structurally interesting. Software innovation is cumulative and modular: each advance builds on previous code, and the ability to read, modify, and redistribute previous code accelerates the accumulation. The exclusivity model imposes costs at every layer of this stack — license fees, incompatibilities, closed interfaces, barriers to modification. Open-source removes these transaction costs. The result is faster iteration and broader contribution.

Firms that initially resisted open source eventually found they could build profitable businesses around it through services, support, complementary proprietary products, and network effects. IBM, Red Hat, Google, Amazon — the largest technology companies in the world built their cloud and infrastructure businesses substantially on open-source foundations. This was not a concession to idealists. It was a recognition that open-source infrastructure reduced their costs and that they could capture value elsewhere in the stack.

Open standards tell a similar story. The internet itself — TCP/IP, HTTP, HTML — was built on open standards developed through the Internet Engineering Task Force and the World Wide Web Consortium, not through proprietary protocols. The alternative — a fragmented internet of incompatible proprietary networks — would have been far less valuable. The openness was not a failure to capture value; it was the creation of value that could not have existed under a proprietary model.

Academic publishing represents a case where the IP model produces outcomes that are difficult to defend in public.

Universities pay researchers to produce knowledge. Governments fund universities through grants and subsidies. Researchers produce papers and submit them to journals for free. Journals send papers to other researchers for peer review, for free. Journals then publish the papers and sell subscriptions back to universities at prices that have risen dramatically faster than inflation — Elsevier's profit margins have consistently exceeded 30 percent. A university library might spend tens of millions of dollars annually to give its researchers access to research that its own researchers produced.

The publishers' argument is that they add value through organization, editorial oversight, and the prestige of specific journals. This argument has weakened as digital distribution has made the distribution function nearly costless. The prestige function is real but circular: journals are prestigious partly because researchers publish in them, and researchers publish in them partly because they are prestigious.

Open-access mandates have begun to revise this arrangement. The NIH's public access policy, requiring that NIH-funded research be deposited in PubMed Central within twelve months of publication, was enacted in 2008. The White House Office of Science and Technology Policy's 2022 memo requiring immediate open access for all federally funded research — without the twelve-month embargo — represented a significant escalation. The European Plan S initiative, adopted by a coalition of research funders, requires that research funded by participating funders be published open access immediately.

The preprint server model — arXiv in physics, bioRxiv and medRxiv in biology and medicine — has in some fields effectively displaced journal publication as the primary mechanism of knowledge sharing, with journals providing post-hoc credentialing rather than primary distribution. During the COVID-19 pandemic, the overwhelming majority of important epidemiological and virological research circulated as preprints before formal publication, because the speed of the crisis demanded it.

The evidence that open access increases knowledge use is robust. Open-access papers receive more citations, including from researchers in lower-income countries that cannot afford subscription access. Practitioners — clinicians, engineers, policymakers — who are not connected to university libraries can access the literature. The cumulative effect on the pace of knowledge diffusion is substantial.

The drug development case is the hardest and most important.

The patent-based pharmaceutical model has produced significant innovation. It has also produced systematic failures: neglect of diseases that primarily affect poor populations, high prices that limit access in wealthy countries and make drugs unavailable in poor ones, gaming of patent systems to extend monopoly periods beyond the original term, and a focus on conditions affecting large and wealthy markets regardless of global disease burden.

The neglected tropical diseases problem is the starkest illustration. Diseases like sleeping sickness, leishmaniasis, and Chagas disease collectively cause millions of deaths and hundreds of millions of disability-adjusted life years. They receive almost no pharmaceutical R&D because the populations they affect cannot pay prices that recoup R&D costs. The patent model, in this domain, does not fail to incentivize innovation; it fails to direct innovation toward where suffering is greatest.

Prize models are the most discussed alternative. Rather than granting a temporary monopoly, a prize model would have governments or international bodies fund R&D through advance market commitments or prizes for successful development of specified medicines, with the successful compound then licensed openly. The innovation incentive is provided; the monopoly pricing is not. The Access to Medicine Foundation, Medicines for Malaria Venture, and Drugs for Neglected Diseases initiative are partial experiments in this direction, funded through philanthropic and government sources.

The COVID-19 pandemic demonstrated both the best and worst of the patent model in pharmaceutical development. The best: mRNA vaccine technology, developed over decades through a combination of academic and private research, was successfully applied to produce effective vaccines with extraordinary speed. The worst: vaccines developed with substantial public funding through Operation Warp Speed and equivalent programs in Europe were patented and allocated primarily to wealthy countries through bilateral deals, producing severe inequity. Waiving vaccine patents, as proposed through the WTO, was blocked by pharmaceutical companies and the governments of wealthy countries, while billions of doses went to waste in wealthy countries and low-income countries remained at single-digit vaccination rates.

The argument that pharmaceutical patents are necessary for innovation does not engage seriously with the degree to which current pharmaceutical R&D is publicly subsidized. NIH funding underlies most of the basic science that produces drug targets. Clinical trials receive public funding. Regulatory approval processes are publicly funded. The marginal contribution of private IP investment is real but takes place against a background of massive public infrastructure that is not captured in the "billion-dollar drug development" figure. Revising the IP framework to more accurately allocate returns proportional to contribution would look quite different from the current system.

The public domain — knowledge and creative work available to everyone, freely — is not a residual or a failure of the IP system. It is the IP system's intended output. Copyright and patents are temporary grants of exclusivity designed to produce work that eventually enters the commons. The public domain is the purpose; the monopoly is the mechanism.

The progressive extension of copyright terms has delayed this output. Works that would have entered the public domain decades ago remain protected. Mickey Mouse, created in 1928, will enter the public domain in 2024 — but only because the Sonny Bono Act extended terms before the 1928 works would have crossed the threshold, and not before Disney had lobbied successfully to prevent it from happening earlier. The economic benefit of copyright to the original creators largely accrues in the first few years after publication; the multi-decade extensions primarily benefit corporate rightsholders who have acquired the rights.

The Creative Commons licensing system, developed in the early 2000s by Lawrence Lessig and colleagues, created a set of standardized licenses that allow creators to grant specific permissions while retaining copyright. A creator can allow non-commercial use, require attribution, allow derivative works, or require that derivative works be licensed under the same terms. This has created a large and growing commons of openly licensed creative work — Wikipedia, Flickr's licensed photographs, open educational resources, open government data — that exists within the framework of copyright law while functionally revising the exclusivity norms that copyright imposes by default.

The revision of intellectual property law toward open knowledge models is not a single policy change. It is a systemic revision across multiple layers.

The funding model for knowledge production must change alongside IP law. Open-access mandates without changes to research funding create perverse outcomes: researchers are required to publish open-access but must pay processing fees to do so, which are affordable for well-funded researchers at wealthy institutions and prohibitive for researchers in the global south and at under-resourced institutions. Open access funded through article processing charges shifts rather than removes the financial barrier.

The international dimension is unavoidable. IP law is now heavily shaped by international trade agreements — TRIPS under the WTO, bilateral free trade agreements pushed by the United States and European Union — that impose high IP protection standards on developing countries as a condition of market access. Countries that might prefer to use compulsory licensing, price controls, or weaker patent regimes face pressure to maintain standards that benefit incumbent rightsholders in wealthy countries. Revising global IP law toward open knowledge requires revising the international political economy that maintains the current standards.

The transition costs are real. Industries that have organized their business models around IP exclusivity — pharmaceuticals, media, software in some domains — would face disruption from revision. This creates powerful constituencies for the status quo. The history of IP law is in part a history of these constituencies successfully defending and expanding their positions.

But the demonstration that open models can outcompete closed ones — that Linux beats proprietary Unix, that Wikipedia defeats Britannica, that open scientific databases accelerate research — means that the revision argument is no longer primarily theoretical. It is empirical. The evidence is in production.

What happens when intellectual property law revises toward open knowledge: more gets made, more gets used, more people can contribute, and the benefits of knowledge production diffuse further and faster. The constraint is not that this is unknown. The constraint is that the people who benefit from the current model are organized, and the people who would benefit from the revision are diffuse.