We are in the midst of a once-in-a-generation opportunity to remake our approach to science. This moment, in all its difficult uncertainty—COVID-19, economic turmoil and the crescendo of a long overdue national discussion about racial justice—demonstrates why universities, funders and other research stakeholders should move decisively to embrace open science. By adopting what are called “open science” practices, we can align the incentive structures of research production and consumption with our values, and catalyze the scientific progress our society so desperately needs.
The two of us—the president of Arizona State University (ASU), which has topped U.S. News & World Report’s “Most Innovative Schools” list since the inception of the category in 2016, and the director of the Open Research Funders Group (a collaboration of leading philanthropies that collectively confer more than $10 billion in grants annually)—call on our peers to commit not only in principle, but also in practice, to creating a more efficient, effective and equitable research ecosystem.
Open science, to quote Michael Nielson’s Reinventing Discovery, is “the idea that scientific knowledge of all kinds should be openly shared as early as is practical in the discovery process.” That open science is an integral tool in the fight against COVID-19 is indisputable: the importance of access to scientific articles and data to help identify promising vaccines and therapeutics was recognized by publishers and researchers alike early in the pandemic. As a consequence, the research community has worked rapidly to take down the barriers—including article paywalls, data hoarding and siloed lab work— that chronically impede scientific progress.
The open dissemination, discussion and testing of COVID-19-related science has quickly taken the place of these outdated norms. Within one month of the first reported case, the virus was rapidly sequenced and openly posted to GenBank, the NIH genetic sequence database. Scores of researchers racing to learn more about COVID-19 shared their early findings as openly accessible preprints. These findings were tested and refined in real-time discussions that were tracked publicly and transparently. Papers that could not withstand replication and reproducibility efforts were quickly and publicly debunked, allowing the scientific community to pursue more promising research avenues. Society and commercial publishers made subscription-controlled coronavirus articles available to all. The protocols and technology behind the Yale School of Public Health’s COVID-19 saliva test have been made available as open source.
Two clear conclusions can be drawn from this rapid alignment. First, the daily workings of science have practical ramifications in all our lives. Scientific norms affect not just researchers working in labs, but also policy makers, doctors, patients, families, and the general public. Second, open science is the form of research dissemination and global collaboration that best reduces vexing limits to knowledge that are exacerbated by COVID-19. If rapidly and openly sharing research data and papers is critical to understanding and combating coronavirus, doesn’t the same hold true for cancer? Heart disease? Climate change? The scientific community—moving with great speed and clarity of purpose—has clearly signaled that open science is the most efficient way to tackle issues that have a significant and direct effect on the lives of the general public. The unambiguous conclusion is that open is better for science.
Importantly, open is also better for the economy. For example, around the turn of the century, the massive and massively successful Human Genome Project placed research results in the public domain. This commitment to open science generated nearly $800 billion dollars in economic benefits between 1988 and 2010, a return on investment of $141 for each dollar of the federal government’s investment in the project. More than 310,000 jobs in the U.S. economy were created, directly and indirectly, totaling almost four million job-years of employment as a result of this scientific undertaking. Similarly, we also owe the development of global positioning systems to the real-world implementation of open science principles, a development that produced more than $50 billion in economic benefits.
Indeed, across a range of sectors from health care to energy, a McKinsey estimate from 2013 put the potential economic value of open data alone in the trillions of dollars annually, equivalent to more than three times the global economic impact of the automotive industry. By leaning into open science practices, we can fuel innovation, job creation and economic growth. As Franklin Roosevelt opined in 1941, at a similar moment of upheaval and uncertainty, “the enjoyment of the fruits of scientific progress in a wider and constantly rising standard of living” is one of America’s basic expectations.
In addition to being better for science and better for the economy, open is better for society. ASU has found strength in defining success not in whom it excludes, but whom it includes. When knowledge and innovation rest in the hands of the few, we struggle to reach our collective potential. Access to data and published research democratizes information and allows more voices to join the scientific conversation. It removes a layer of insularity in ways both big and small. To take one example at the systemic level, the average library expenditures at Historically Black Colleges and Universities (HBCUs) are significantly less than those of non-HBCU counterparts. This translates in real terms to a racialized inequality of access to the journal articles, books, and other materials upon which future research can be formulated.
At the individual level, the exchange of scientific information often occurs in direct personal interactions. Data that are otherwise proprietary may be shared among close peers and colleagues. Scholars without access to paywalled articles can request copies from the authors, but may be hesitant to do so if they are not part of the same informal networks. By making these materials open for all—to access, replicate, question and build upon—we can contribute to both levelling the playing field and widening the circle of science.
Universities, philanthropies, government agencies and other stakeholders can accelerate the positive effects of open science—in the fight against COVID-19, in our efforts to strengthen the economy, and in our quest for a more just society—by aligning our incentive structures with our values. Practically, this means exercising specific points of leverage—including hiring, funding, tenure and promotion—to ensure that research practices become more open. Many examples are flowering today. Dozens of university departments include language in their job postings along the lines of “This department values transparent, replicable research and open science principles.” This sets the expectation that open practices will be a component of not only the job interview but, for the successful candidate, of the job itself. A wide range of philanthropies are now asking grant applicants to explain how they have historically made their work open, and how, if funded, they will make their outputs open going forward. This provides a powerful incentive (the promise of financial support) for researchers to adhere to open practices.
Aligning research incentives to reward open science practices may seem daunting, but university and philanthropic leadership can start the process by taking specific, concrete actions that have already been proven effective in practice. While a number of organizations have launched fully actualized open science programs, notably McGill’s Montreal Neurological Institute-Hospital (The Neuro) and the Rochester Institute of Technology’s Open@RIT, university presidents and provosts can move their institutions systematically toward open simply by engaging in a structured dialog with their researchers. In this spirit, we call on universities to emulate MIT and launch an open science task force. MIT’s work began with a concise charge from its provost, to “coordinate a renewed Institute-wide discussion of ways in which current policies and practices might be updated or revised to further the Institute’s mission of disseminating the fruits of its research and scholarship as widely as possible.”
The MIT model is a true collaboration among the administration, chairs and faculty that includes the development and deployment of open science plans tailored to the disciplinary considerations of each department. It is predicated on the acknowledgement that what constitutes open science best practices in, say, anthropology, will differ from what works for zoology.
Facilitating this bespoke departmental approach are the many emerging norms and policies articulated by professional societies such as the Linguistics Society of America, the Society for the Improvement of Psychological Science and the American Geophysical Union. Ideally, the recommended policies that arise from these task forces will resonate with faculty from both an institutional and a disciplinary perspective. The ultimate goal is to develop "mutually reinforcing vectors," an environment in which researchers hear consistently from a range of influencers at their university, within their discipline, and across potential funding sources that open practices are both warmly encouraged and properly rewarded.
In support of this "mutually reinforcing vectors" approach, we also call on philanthropies to adopt grantmaking policies that encourage researchers to share their outputs (articles, data, code, materials, etc.) openly and rapidly. In this effort, they can lean on the work of funders ranging from the American Heart Association to the Gates Foundation to the Michael J. Fox Foundation to craft language and workflows that have been field-tested over thousands of grant conferrals. Philanthropies can also draw from policy language templates (developed by the Open Research Funders Group and endorsed by funders including the Sloan Foundation and the Wellcome Trust) to implement a stepwise approach to more closely aligning their incentivization schemes with open science principles.
While open is better for science, the economy, society, it is not magic. It takes concerted, direct effort by key stakeholders to effect change. It also takes a community of practice—sharing successes, roadblocks, and solutions; developing and testing resources that explain the whats, whys and hows of open; and identifying key opportunities to expand the “coalition of the willing.” One such effort is the National Academy of Science, Engineering, and Medicine’s Roundtable on Aligning Incentives for Open Science (in which we both participate). The Roundtable includes direct representation from colleges and universities, philanthropies and government agencies.
Crucially, the broader network of stakeholders engaging with the Roundtable also includes more than 500 professors, postdocs, librarians, professional society representatives, publishers, funders and other stakeholders. For any university or philanthropy finding itself not yet prepared to take the plunge in the manner we have outlined above, we warmly encourage you to engage with the roundtable to get a better sense of the tangible steps your peers are taking to stimulate open science within their institutions.
There are hurdles to widespread adoption of open science practices, to be sure. Researchers need proper training on data management plans, reuse licensing and other good open science hygiene. Infrastructure must be developed and nurtured to preserve scientific data, curate it and render it actionable. And organizations must overcome their natural entropy, which makes tackling big, cross-cutting initiatives like open science challenging. While these obstacles are nontrivial, they are small in comparison to the scientific, economic, and societal benefits of open. In a moment of great peril, maintaining the status quo will ultimately prove more costly.