What happened
Two years into a ten-year global framework for science, UNESCO published its first scorecard on whether the thing is working, and then spent ninety minutes on stage arguing that the scorecard is the wrong way to judge it. That tension is the whole session, and it is unusually candid for a launch event.
The first Global Report of the International Decade of Sciences for Sustainable Development is built on 397 endorsed scientific initiatives across 79 countries, supporting all 17 Goals, with USD 50 million in confirmed funding mobilised. Its two sharpest findings are about distribution rather than volume: Africa accounts for under 10% of the initiatives while holding 17.5% of the world’s population, and 40% of the initiatives named coordination, not funding, as their primary barrier. UNESCO frames the Decade’s core problem the same way its Director-General did.
Presenting the report, UNESCO’s Director of Basic Sciences, Engineering, Technology and Innovation put the diagnosis in one line: the main barrier is “the failure to translate the scientific knowledge into policy, governance and action and scale.” He then said the quiet part about his own numbers, and it is the quote to keep.
What the recording actually covers
The published programme for the morning listed three items: the report launch from 09:30 to 11:00, a coffee break and media briefing from 11:00 to 11:30, and parallel thematic sessions from 11:30 to 13:00. The recording follows none of that. It opens on the chair welcoming the room to “the second day” of the conference and the launch of the report, runs the launch panel to roughly the 1:26 mark, breaks for about seven minutes while the stage is reset, and then runs a second plenary panel, Translating evidence into priorities, to the end. It closes on the chair announcing lunch and saying that “at two p m” there will be thematic events and side activities, with the room to reconvene at 14:00.
So the morning ran as two back-to-back plenaries into a lunch break. There is no coffee break, no media briefing and no thematic session in this recording, and the thematic sessions appear to have moved to the afternoon. Two other things billed did not happen: the Q&A the chair announced at the top was cut for time (“we were supposed actually to have questions from the audience, but as you have seen the discussion was so interesting that we couldn’t uh really have enough time”), and the QR code meant to put the report on the screen never appeared, leaving the moderator to point the room at the screens in the lobby instead. For a launch whose central complaint is coordination, both are small ironies worth recording.
397 is a count of registrations, not of results
The Decade runs on what UNESCO calls a decentralised mechanism: it issues calls, anyone from a research consortium to a single person can propose an initiative, and UNESCO evaluates and endorses. The 397 is therefore a count of things people submitted and UNESCO accepted. It measures participation. It does not measure whether a single SDG target moved.
To the launch’s credit, nobody pretended otherwise. The presenter’s own list of what needs strengthening included “strengthen in the monitoring and accountability,” with the reasoning stated plainly: the Decade already has many endorsed initiatives, but it needs to see how they are implemented. The panellist from the United Nations University made the same point about the wider system, and made it best.
She extended it into a rule for the Decade itself: success “is not measured by the number of articles, the number of reports we churn out,” and the Decade “should not be measured by the science or the amount of science generated, but by whether it’s been able to have better policies generated, stronger institutions developed, and tangible progress for people on the ground.” That is three separate senior people at a report launch saying the report’s headline number is not the thing to watch.
The number that says most about the Decade is not 397. It is 29. The presentation reported that 29 member states have nominated a focal point to the Decade, and did not dress it up: the number “is not usually what we want. We want more members to be committed, to be engaged in this process.” A UN decade proclaimed by the General Assembly has, two years in, a named counterpart in fewer than one in six of the states that proclaimed it. Initiatives can be counted because scientists volunteer them. Governments are the harder count, and they are the count that is short.
The equity finding was named, and then softened
Africa under 10% of initiatives against 17.5% of world population is the report’s hardest number. The launch did not bury it, but it did not dwell on it either. The presentation put it as a matter of uneven participation, noting most initiatives come from Asia-Pacific and Europe and that on Africa and the Arab states “we don’t have much very impressive” numbers, and it pointed at the structural cause behind it: the United States, China and the European Union account for roughly “two third of the global investment in R and D,” with researcher density and research infrastructure distributed accordingly.
The sharpest response came from the International Science Council, whose Science Director declined the easy remedy.
Her alternative was to set “ambitious and maybe even audacious targets for investment” that build national science systems, which is a demand for a number that does not currently exist in the Decade’s reporting. McBride also read across two of the report’s findings to locate the gap precisely: a high response on SDG 9 (innovation) against a “quite low response on SDG sixteen on institutions,” which she read as “a kind of gap between innovation and governance” that is now widening. The presentation reported the same skew, with initiatives clustering on education and innovation rather than hunger, poverty, food and ocean, and gender equality “still not get the proper attention.” The Decade, in other words, is best represented exactly where science is already strong and worst represented where the Goals are furthest behind.
Coordination, not money: the finding that inverts the usual ask
The most interesting number in the report is that 40% of endorsed initiatives named coordination as their primary barrier. The moderator gave the other half of that split from the chair: only 10% named funding. That is close to the reverse of what a UN science programme is normally told, and the presentation treated it as a symptom rather than a complaint, saying the initiatives’ own reporting reflects “something structural” in the global science system.
The panel largely converged on what coordination machinery would look like: permanent science-policy interfaces instead of one-off consultations, trusted institutions acting as knowledge brokers, and long-term rather than project-cycle funding. Michel Spiro argued for thousands of transdisciplinary hubs pairing academics with citizens, NGOs, the private sector and decision-makers, and offered CERN’s open Zenodo repository as somewhere to deposit what they produce. But the answer with an accountability mechanism attached came in the second panel, from South Africa’s Deputy Minister of Science, Technology and Innovation.
Budgets and procurement is the operative part. It is the only formulation offered all morning that would leave a trace in a document someone audits, and it came from a country that also put its own number on the table without being asked: South Africa spends 0.62% of GDP on R&D, against unemployment above 30%.
The second panel was where the abstraction ran out generally. Asked for one case where evidence actually changed policy, the vice-president of Uzbekistan’s Academy of Sciences, a former deputy minister, told the Aral Sea story: scientists identified salt-tolerant plants for the dried seabed, the government funded a national programme, more than a million hectares were planted over five years, monitoring showed soil recovering and dust storms falling, and three presidential decrees followed. Her diagnosis of why evidence usually fails is the cleanest sentence in the session: “scientists sit in one room” setting priorities while “government sit on another room” writing theirs. A former director of UNESCO’s regional bureau for sciences in Cairo, now an independent consultant, went one step further back in the chain: “To establish evidence, you need data.” He described a UNDP-era knowledge report on which every country surveyed failed the threshold, and one government, Morocco, that responded not by ignoring it but by spending three weeks examining the method step by step with his team, then accepting the result and disseminating it. That is what accountability looks like when it works, and it took three weeks of adversarial review to get there.
The open science half of the count
UNESCO gave a second publication, the Open Science Outlook, its own launch in a parallel session that same afternoon, and its figures are the ones that should worry anyone who tracks follow-through. 81 countries have implemented UNESCO’s 2021 Recommendation on Open Science. Of governments, 79% report an open science policy framework, 41% have an implementation plan, and only one in three has a monitoring mechanism.
Read as a funnel, that is a commitment surviving into a plan roughly half the time and into measurement roughly a third of the time. Policy outruns the machinery that would check it by about two to one. The same shape runs through HLPF a week earlier, where governments arrived able to report more SDG indicators than ever and their own civil society kept asking what the improved averages still could not see. Two of the Decade’s own priorities point at the same joint from the science side: “reform the research assessment system at the global national levels,” because a system that rewards publication over local problems will keep producing the skew the report just measured, and, from the Indian panellist, a request that UNESCO intervene on publishing costs, which have made scholarly publication “a big business” that researchers struggle to afford.
Our read
A stocktake is a counting exercise, and this one counted the thing it could count. 397 endorsed initiatives is a real and fast-assembled number, and UNESCO is right that it shows a mobilised science community. But it is a register, not a result. It records who raised a hand. The Decade has no published measure of whether any of the 397 changed a policy, and the man who presented it said so from the podium.
Put the three numbers in a row and the Decade’s real position appears. 397 initiatives, 79 countries, 29 focal points. Participation is broad, scientific enthusiasm is not the constraint, and government ownership is thin. That is the same finding the initiatives themselves reported when 40% named coordination rather than funding: the missing input is not money or knowledge, it is the institutional wiring that turns either into a decision. The Open Science Outlook then puts a ratio on the wiring, one in three governments with a monitoring mechanism against 79% with a policy.
The honest version of the tension is this. The launch diagnosed itself accurately, in public, in more candid terms than most UN reports manage. It did not commit to a number that would let anyone check the diagnosis next time. The panel asked repeatedly for permanent science-policy interfaces, and exactly one speaker named the artefacts that would prove one existed: budgets and procurement. The next Global Report will be able to say whether 397 became 600. Whether it can say what any of them changed is a decision UNESCO has not yet made, and it is the only decision that would make the second stocktake worth more than the first.
Why it matters for the SDGs
The Decade exists because SDG 17 promised science, technology and capacity-building as means of implementation for everything else, and this report is the first attempt to score that promise. Its findings map onto the Goals unevenly and revealingly: strong on SDG 9 (innovation), weak on SDG 16 (institutions), thin on hunger, poverty and ocean, and short on SDG 5 (gender equality), which the presentation said is still not getting proper attention. The Africa figure bears directly on leaving no one behind, since a Decade whose initiatives cluster where research capacity already sits will tend to widen the divide it was meant to close.
There is also a measurement argument here that connects to the wider July agenda. Michel Spiro used his intervention to press for progress indicators other than GDP, naming the Human Development Index, the Club of Rome’s genuine progress indicator and Bhutan’s gross national happiness index, against a GDP that he called “the only index which is put forward nowadays.” That is the same argument HLPF held a week earlier in Beyond GDP, arriving from the science side rather than the statistics side. The Decade’s own answer to it, reforming how science is assessed so that solving a local problem counts for as much as a paper, is a Beyond GDP argument aimed at researchers instead of finance ministries.
Watch & read
- The first IDSSD Global Report, UNESCO (UNESDOC), the document launched at this session.
- “Persistent divides on science threaten to leave developing nations behind”, UNESCO’s own summary of the findings, published 15 July and the source for every published figure on this page.
- The conference and the Science Decade programme site.
- The measurement argument next door: Beyond GDP at HLPF 2026, and the full HLPF coverage, where the same commitment-outruns-measurement pattern runs through the Voluntary National Reviews.
- Full IDSSD Global Conference 2026 coverage.
Quotations are lightly edited from an automated transcript of the UNESCO webcast and should be read as close paraphrase. Speaker labels are not reliable across this recording, so speakers are named only where the content makes the attribution unambiguous and are otherwise cited by role. We have not read either report, so their figures are as UNESCO published them in its own summary, and each is attributed to the publication that carries it: the initiative, funding, Africa and coordination figures are from the IDSSD Global Report; the open science figures are from the separate Open Science Outlook. Figures stated at the launch that UNESCO’s published summary does not carry are marked as such in the text. UNESCO published its summary on 15 July; the launch session covered here ran on the morning of 16 July. Non-English remarks are summarised, not quoted.