Open science has a monitoring problem. In the years since UNESCO’s Recommendation on Open Science was adopted on 23 November 2021, the biggest headlines have gone to publisher negotiations, the growth of preprint servers and generative AI’s incursion into literature search. The quieter event is a questionnaire. Member States are expected to report, every four years, on the policies, sustained funding and physical infrastructure they have actually put in place. That exercise has moved from abstract commitment to national paperwork, and the first reports are the most honest test the Recommendation has faced.
The date matters because the instrument is standard-setting, not a treaty. It carries no fines and no legal penalties. Whatever force it has comes from visibility: governments must explain themselves, in public, against a shared international yardstick. That is less dramatic than a signing ceremony. It may be more useful.
What the Recommendation Actually Asks
UNESCO’s General Conference had asked for a standard-setting instrument in 2019, and negotiators spent two years settling the definition. The final text covers open scientific knowledge, including publications, research data, software and hardware, and it extends to open science infrastructures, open engagement of societal actors and explicit dialogue with other knowledge systems, including Indigenous and local knowledge. Member States are invited to translate that definition into national law and practice. The reporting requirement sits inside this framework. Each country is expected to describe its open science policies, its investments in repositories and data services, its training and capacity-building work, and any changes to how researchers are evaluated.
The shift is subtle but real. Older declarations asked governments to support open science. The Recommendation asks them to document what they support, with whom, at what cost, and under which authority. For many research ministries, that is the first time anyone has asked those questions together.
Why the First Reporting Cycle Matters
The baseline was not a state filing. In 2023, UNESCO published the first Open Science Outlook, a global scan that drew on surveys and regional analysis rather than formal government reports. It gave a partial picture. The new member state reports are meant to be the official record: country-level claims, received under the Recommendation’s own implementation process, comparable across regions and across years.
That comparison is where the leverage sits. A national report is not just a document for UNESCO; it is a public reference point for researchers, university consortia and donors. If a government says it has a national open science policy, someone can check whether repositories are funded through a real budget line. If it says assessment reform has begun, someone can watch whether hiring committees still default to journal brand names. The UNESCO open science programme has developed a monitoring framework with indicators for policy, infrastructure, capacity and engagement. The reports are supposed to populate those indicators with national evidence, turning a list of good intentions into a comparable dataset.
None of that happens automatically. It depends on ministries collecting numbers they rarely collect, and on universities supplying numbers they sometimes prefer not to share.
Photo by National Cancer Institute on Unsplash
Here’s the Catch
Here’s the catch: the mechanism is only as strong as the data standards behind it. UNESCO can ask, but it cannot compel. There are no sanctions for a late, thin or demonstrably self-flattering submission. In some countries, responsibility for open science sits with a research ministry, in others with an education ministry or a science council, and in a few nowhere that can be identified on an org chart. Fragmentation shows up directly in report quality.
Self-reporting creates a second problem. Countries are being asked to grade their own homework. A government can list a national open science policy that exists on paper while repositories remain underfunded and data-sharing rules go unenforced. Unless the reports are checked against independent sources, the first cycle may measure policy production rather than scientific practice. The Recommendation on Open Science itself acknowledges that implementation depends on national context, and that flexibility is both a strength and an escape hatch.
The risk is not that countries lie. The risk is that everyone reports a plan and nobody reports a budget. A plan is free. A funded repository is not.
Reality Check: Hype Versus Verification
Reality check. The Recommendation’s ambition is high, and the monitoring design is genuinely useful. The verification side is weak. Grade the ambition A-minus. Grade the enforcement architecture D-plus. That does not make the reporting exercise meaningless. Comparisons are powerful. A country that reports no national repository policy while its neighbours report funded data infrastructure will feel the gap.
The real value is not the report. It is the preparation. Ministries that had no open science working group now have to convene one. Universities that never counted APC spending now have to find the invoices. Data officers who had no standing now have a document asking questions they have been trying to get answered for years. That preparation changes institutions, even when the final text is bland.
What Universities and Research Offices Should Expect
Research universities are the primary data sources for this process. Institutional repositories, data management plans, APC payments, research information systems and policy registers are what national reports ask about. A university that cannot say how many datasets it preserves, at what cost, and under which licence is not ready to answer the national questionnaire. That gap will surface quickly.
National policy already moves faster than institutional habit in several places. India’s One Nation One Subscription deal changed how universities pay for journal access, and Japan’s national open access policy pushed institutional repositories from optional projects to expected infrastructure. Both are exactly the kind of state action the UNESCO reporting exercise is designed to make visible. For an early-career researcher, those reports are not abstractions. An institution with a funded repository and a data-sharing policy is a different place to work from one where open science means a PDF on a website.
What to Watch After the First Reports Land
The reports will be most useful when read side by side, not one country at a time. Watch for countries that publish the underlying questionnaire and supporting evidence, not just a summary. Watch for whether repository usage statistics are current or years out of date. Watch for whether APC spending is broken down by publisher or lumped into a single procurement line.
Those details matter because they show whether a government is willing to be examined at the level where open science actually lives: the institutional budget, the metadata export, the data steward’s contract. A report that avoids those details has answered the wrong questions.
The Last Word
Ask a researcher who has spent an afternoon formatting a data deposit for a repository that will not accept it, and the Recommendation looks less like a global framework and more like a set of promises. The reports will matter only if they name budgets for repositories, metadata export standards that actually work across systems and staffing for research data support. A country can write “open science is a priority” in a report. A researcher can tell the difference before lunch.
The last word belongs to the person who has to deposit the data, not the publisher who sells the access. If the first cycle produces comparable public accounts of who funds open infrastructure and who does not, it will have done more for open science than most of the announcements that preceded it.