The impact of COVID-19 on the debate on open science: a qualitative analysis of published materials from the period of the pandemic

Open science (OS) and open access (OA) have been debated for over two decades and increasingly positioned as integral to a well-functioning research system. The COVID-19 pandemic acted as a major test of the global research system, foregrounding issues of research productivity, equity, diversity and inclusion (EDI), and the system’s capacity to respond to emergencies. During the pandemic, advocates argued OS demonstrated clear benefits—some claiming “open science saves lives”—while skeptics warned that certain OS practices risked disseminating unvetted science and fueling misinformation. This study asks: How did the COVID-19 pandemic impact the debate on open science? It qualitatively analyses a multilingual corpus of articles, editorials, blogs, and thought pieces published during the pandemic to examine shifts in arguments, focal topics, and framings of OS. The importance of the study lies in clarifying whether and how the crisis reconfigured OS discourse, sharpened lines of sight between OS and societal benefit, and influenced future policy and practice trajectories.

Background debates on OS and OA predate COVID-19, encompassing open data sharing, open peer review, and research assessment reform, alongside concerns about sustainability, business models, incentives, and global equity. Previous health crises (Ebola 2014–16; Zika 2015–16) spurred calls for rapid and responsible data sharing and preprinting in emergencies, culminating in coordinated openness statements by funders and publishers (Wellcome 2016; 2020). During COVID-19, publishers widely opened COVID-related content, and major policy initiatives reinforced OS, notably the 2022 US OSTP (Nelson) memo mandating immediate public access to federally funded research and UNESCO’s 2021 Recommendation on Open Science, which broadened OS to include engagement with societal actors and dialogue with other knowledge systems. Pre-pandemic literature highlighted disciplinary and regional variations in OS uptake and persistent equity concerns (e.g., Global North/South imbalances, language barriers), all of which provided a backdrop against which the pandemic-era debate unfolded.

Design: Qualitative inductive content analysis using thematic analysis. Corpus construction: Initial queries in the Open Access Tracking Project combining OS terms (e.g., open access, preprints, open data) and COVID-19 terms (e.g., coronavirus, pandemic) across English, German, Portuguese, and Spanish for Dec 2019–Dec 2022. Results filtered by date, language, and source type, then manually screened for relevance. Augmentation: Additional searches in Web of Science and Dimensions (peer-reviewed articles, editorials), Dimensions (conference proceedings, grey literature), BASE (peer-reviewed and grey), Overton (policy documents), Nexis (media/press releases); targeted searches of key policy/professional websites (UNESCO, UN, WHO, Science Europe) and language-specific databases (e.g., SciELO, idw). Items assembled in Zotero, deduplicated, and screened. Final corpus: 446 items total—311 English, 53 German, 31 Portuguese, 51 Spanish. By type: 145 newspaper/magazine/press, 141 journal articles, 103 blog/opinion, 36 reports/policy, 8 conference papers/presentations, 8 preprints, 5 books/chapters. Analysis: NVivo 1.7.1 used for coding. A core set of 32 English pieces was deeply open-coded to develop a 128-code codebook, validated by team members. Remaining English corpus coded via memos mapped to codes. Parallel coding in German, Portuguese, and Spanish by fluent team members using and adapting the codebook; two additional codes added (total 130). Themes were developed from coded data and reported in Findings and Discussion.

- Pandemic as ‘stress test’ and accelerator: The pandemic was widely framed as a ‘stress test’, ‘catalyst’, ‘revolution’, or ‘tipping point’ for OS, showing that rapid, open collaboration is feasible with sufficient political will. Some authors reported changed views on OS due to observed benefits during COVID-19. - Clear line of sight to societal benefits: Many argued OS accelerated scientific discovery and policy responses, leading to tangible health outcomes (diagnostics, treatments, vaccines) and broader societal benefits (economic gains, evidence-informed policy, public communication). Arguments were largely pre-existing but gained urgency and concreteness in the pandemic context. - Instrumental benefits to science: OS was said to speed dissemination and knowledge production, reduce waste and duplication, improve transparency and accountability, foster collaboration and interdisciplinarity, and potentially enhance research quality and integrity. These points were amplified by pandemic needs for speed and coordination. - Shifts in debate focus: Discussion of OA business models and APCs receded as publishers temporarily removed COVID paywalls. Instead, debates concentrated on open data sharing (OD), preprinting, information quality, retractions, and misinformation. - Open data sharing: Early successes (e.g., genomic surveillance, vaccine development) highlighted OD’s value, but concerns emerged about uneven uptake, infrastructure gaps, coordination, and interoperability. In German-language sources, OD debates emphasized government data sharing for efficiency, innovation, competitiveness, and public trust, reflecting regional policy priorities. - Preprinting: Use and acceptance of preprints rose notably in early 2020, with major journals signaling acceptance of preprinted work. Advocates emphasized speed and community feedback; critics warned of ‘unvetted science’ and media misuse. Some high-profile problematic preprints were rapidly corrected/withdrawn, illustrating post-publication peer review; a preprint withdrawal rate of 0.26% was cited. - Quality, retractions, misinformation: The tension between speed and rigor intensified. High-profile retractions in leading journals (e.g., Surgisphere-based studies) showed that closed, peer-reviewed systems are also fallible. Risks of an ‘infodemic’, persistent citation of retracted work, and challenges of fully withdrawing flawed research were highlighted, underscoring journalists’ role in contextualizing science. - Reframing OS: Moves to conceptualize OS as an integrated whole (OA, OD, open peer review, open metrics/assessment, preregistration) and to broaden scope (UNESCO’s pillars: open scientific knowledge, infrastructures, engagement with societal actors, dialogue with other knowledge systems). While engagement gained traction, dialogue with other knowledge systems remained peripheral. - Equity and critical perspectives: Critical discussions, especially Global North/South dynamics, multilingualism, and Indigenous knowledges, were less prominent than might be expected. The corpus focused largely on Global North experiences, despite multilingual coverage. - Sustainability concerns: Authors worried that openness gains (e.g., temporary paywall removal) might not persist post-pandemic; calls were made to institutionalize ad hoc measures to preserve progress. Data points: - Corpus size and composition: 446 items (311 EN, 53 DE, 31 PT, 51 ES); types: 145 news/press, 141 journal articles, 103 blogs/op-eds, 36 reports/policy, 8 conference, 8 preprints, 5 books/chapters. - Cited preprint withdrawal rate: 0.26%.

The study addresses how COVID-19 impacted the debate on OS by showing that the crisis sharpened and accelerated long-standing pro-OS arguments while shifting attention toward practices most directly tied to emergency response—open data sharing and preprinting—and to questions of information quality and public communication. The pandemic provided a clear illustrative case connecting OS practices to societal benefit (particularly health outcomes) and to instrumental scientific gains (speed, efficiency, transparency), thereby strengthening advocacy claims and influencing policy (e.g., OSTP’s Nelson memo). At the same time, heightened concerns about misinformation and quality revealed risks and limitations, emphasizing the need for robust infrastructures, clearer communication channels, and responsible media practices. The debate also moved toward integrative and broader framings of OS, in line with UNESCO’s recommendation, though engagement with non-academic knowledge systems and Global South perspectives remained limited. Overall, the findings suggest that COVID-19 intensified the urgency and visibility of OS, catalyzing policy and cultural shifts, but also exposed enduring gaps in equity, infrastructure, and governance that must be addressed for durable, globally inclusive OS.

The pandemic period intensified focus on openness and its linkage to societal impact, with many treating COVID-19 as a ‘test case’ demonstrating OS’s benefits. Discourse concentrated on open scientific knowledge and infrastructures (notably data sharing and preprinting), while broader UNESCO pillars (engagement with societal actors, dialogue with other knowledge systems) saw uneven uptake. Advocacy arguments were not new but gained urgency and concrete exemplars, enabling stronger policy momentum (e.g., OSTP 2022). Skeptical perspectives clustered around quality, retractions, and misinformation, underscoring the need for safeguards. Contributions: The study maps how the pandemic reconfigured OS debates, identifying shifts in focal topics, intensification of advocacy claims, policy uptake, and areas of under-addressed equity. Future directions: Institutionalize pandemic-driven open practices beyond crises; strengthen data and preprint infrastructures, curation, and quality assurance; develop guidance and training for journalists and policymakers engaging with preprints and open datasets; expand OS frameworks to meaningfully include multilingualism, Global South perspectives, and dialogue with other knowledge systems; evaluate long-term sustainability and impacts of post-2022 policy changes on OS practices across disciplines and regions.

- Corpus coverage is broad but not exhaustive; selection focused on Dec 2019–Dec 2022 materials and four languages (English, German, Portuguese, Spanish), potentially omitting relevant debates in other languages/regions. - Reliance on published/editorialized materials (news, blogs, policy, scholarly articles) may overrepresent certain voices (Global North) and underrepresent perspectives from Low-Income countries, Indigenous communities, and non-English contexts. - Many items are short-form, leading to broad-brush or anecdotal claims with limited empirical backing. - Gender and other EDI dimensions received little explicit treatment within the corpus, despite independent evidence of pandemic-related inequities. - Qualitative coding, while systematic, involves interpretive judgments; findings describe debate dynamics rather than causal effects. - Generalizability beyond the studied period and into the post-pandemic era remains uncertain.