The UN doesn't have a mandate problem. It has an off-switch problem.
UN80 is the first serious attempt at mandate reform in twenty years. Whether it works will depend on evidence that nobody has yet joined up.
Since 1946, the General Assembly, Security Council and ECOSOC have issued mandates through more than 40,000 resolutions, decisions and presidential statements. More than 85% of active mandates contain no instruction to review them or bring them to an end.
That is the problem in one line. The UN is not short of instructions. It is short of a way to switch them off.
These are not my numbers. Almost every figure in this piece comes from the UN's own 2025 mandate implementation review (A/80/318) or from official budget documents. I have spent 18 years assessing projects for the World Bank, IFC and EBRD, and what struck me reading them was how familiar the pattern is. Activity is well documented. Whether it achieved anything is much harder to see.
We have been here before
In 2006, Kofi Annan handed member states an online registry of mandates and an analytical framework, and told them the review was theirs to carry out. The Secretariat would supply the tools.
That review named four problems:
Burdensome reporting
Overlap between and within UN bodies
Duplicative delivery structures
Mandates without the resources to deliver them
Twenty years later, the 2025 review found all four still there, and worse. Its own verdict is that the 2006 exercise did not deliver what was intended.
It also explains why. The 2006 approach examined mandates one by one, which the UN now describes as an ultimately unmanageable process. UN80 takes a systemic view instead. That is the right call, but it raises an obvious question: what is different this time?
A system that only grows
The instructions themselves are getting longer. Since 2020, General Assembly resolutions have grown 55% in word count and ECOSOC's by more than 95%. In 2024 the average resolution from the three main organs ran past 3,000 words, nearly triple its length thirty years ago. Longer instructions are harder to deliver and harder to measure.
Then there are three words worth watching in any resolution: "within existing resources". In 2024, more than 15% of GA resolutions added new work on that basis, over four times the rate in 2000. New tasks arrive with no new money, so something else is quietly squeezed. Nobody tracks what.
The paperwork grows too. In 2024 the Secretariat published more than 1,100 reports, averaging around 11,300 words each. Nearly 65% were downloaded fewer than 2,000 times. Direct costs of meetings and reports came to more than $360 million, over 10% of the regular budget.
None of this is anyone's bad intent. It is what happens when a system can add work easily and has no routine way to remove it.
Duplication, drift and habit
In the 2026 budget proposals, Secretariat entities cited nearly 4,000 mandate sources to justify their work. In the development pillar, about half of those mandates were cited by more than one entity.
Around 50 citations pointed to mandates that appear to be inactive, including the Millennium Development Goals. The MDGs ended in 2015.
The agenda itself also recycles. More than 30% of the subjects of GA resolutions in 1990 were still the subject of resolutions in 2024. Some problems genuinely endure. Some resolutions are simply habit. Telling the two apart is a question of evidence, not opinion.
Less money, same instructions
The 2026 regular budget, approved in December 2025, came to $3.45 billion. It largely reflects the roughly 15% resource cut and nearly 19% staffing cut proposed under UN80. The budget shrank. The stock of mandates did not. As one Fifth Committee analyst put it, the cut reduces the money available to deliver the same portfolio of mandates.
The proposed 2027 budget is $3.302 billion with 11,150 posts. How much that changes from 2026 depends on the baseline you pick. UN coverage reports a 1.2% reduction. Another analysis puts it 4.3% below the 2026 appropriation. The ACABQ Chair said total requirements would reach $3.6 billion, 4.5% above 2026. None of that is spin. It is three different baselines, and it is exactly why consistent, independent analysis matters.
The regular budget is also only about 5% of UN system resources. In 2023, around 80% of system funding was voluntary, and more than 85% of that was earmarked. More than 60% of grant transactions were under $1 million, and less than 10% went through pooled funds against a 30% target. The money arrives in thousands of small, ring-fenced pieces, and impact gets fragmented the same way.
Tracing money to results is harder still. About 40% of UN system entities have strategic plans, about 30% have integrated results and resource frameworks, and about 60% lack key parts of a results chain. You cannot judge a mandate if you cannot follow its money to what it delivered.
Plenty of tools, no connections
It would be unfair to say the UN lacks transparency tools. It has many, and some are good:
Mandates: the Secretariat Mandate Registry built under UN80, OHCHR's map of Human Rights Council mandates, and UNFCCC's document and decision database, each covering its own bodies.
Money: open.un.org, which draws system-wide finances from audited financial statements, plus the IATI portals many agencies run. UNDP opened data on 6,000 projects as far back as 2012.
Results and records: results.un.org for programme budget results, the UN Digital Library, searchable meeting transcripts and a document comparison viewer.
Each answers one question well. None connects them: mandate to entity to money to result, tracked over time. That connection is where accountability actually lives.
There is a second limit. Nearly every one of these tools is built and fed by the system it describes. That is not a criticism, it is a design constraint. Self-reported data answers "what did we do?" Independent analysis can ask "was it worth it, and is anyone else doing the same thing?"
This is the gap I know best from development finance. Targets, outputs and outcomes are mostly self-reported. Independent checks that KPIs were actually met are rare, and when they exist they are rarely comparable.
Why the next six months matter
Under GA resolution 80/251, a member-state working group is developing criteria to renew, merge or retire mandates. The Secretariat is due to review inactive, duplicative and fully implemented mandates by the end of December 2026. The working group concludes by 30 April 2027 at the latest.
The co-chairs' September progress report was clear that this is an ongoing process, not a finished outcome. The criteria are still being shaped. That makes this the window in which better evidence can still change what happens.
There is also data the UN has never had before. The new Mandate Registry lists 43,078 documents. In the 2027 budget cycle, 2,795 of them are cited by 107 entities. The Pact for the Future alone is cited by 27.
What it would take to see the whole system
The chain runs: resolution → mandate → entity → programme → budget line → deliverables → reported results → evaluation. Every link exists somewhere in public documents. Almost none of them are connected.
Even the small part is a maze. The regular budget is split into hundreds of programmes and subprogrammes, some with as few as three staff.
The hard parts are real:
"Mandate" means a document in some places and a specific instruction in others.
Entities use different results frameworks.
Voluntary funding sits outside the regular budget.
Decades of records exist only as scanned PDFs.
None of these is impossible. Together, they explain why it has not been done.
In 2006, mapping mandates at this level was a staffing problem. With today's AI tooling for document extraction and analysis, it has become a data problem. Do it across years rather than for a single budget, and you get something new: not a snapshot, but a model of how mandates accumulate, migrate between entities and outlive their purpose. In effect, a digital twin of the system.
Systemic reform, done properly, still needs mandate-level evidence underneath it.
The real risk
UN80 is the first serious mandate reform since 2006, and it starts with better data than the UN has ever had. The risk is not bad intentions. It is that reform ends up measured by documents produced rather than results delivered, which is exactly how the last attempt ran aground.
Independent eyes help. That is the space we are exploring, building on years of work analysing IFI project documents at scale.
I would like to hear from anyone who has worked in or with the UN system. Which mandate has outlived its purpose, and why has it not been retired? And if you could ask one question of the 2,795 mandates cited in the 2027 budget, what would it be?
Sources
UN, Mandate implementation review, A/80/318 (2025): mandate stock, review gap, 2006 findings, resolution length, "within existing resources", reports and costs, duplication, inactive mandates, funding mix, results chains, existing registries
UN press release SG/SM/10398 (2006): Annan's mandate review remarks
open.un.org: system-wide financial data source
IISD SDG Knowledge Hub: UNDP open data launch, 2012
UN News, December 2025: 2026 budget
Casque Bleu, Fifth Committee post-mortem: same mandates, less money
UN press release SG/SM/23331 (2026): proposed 2027 budget
Fifth Committee, 5th plenary meeting, 81st session: ACABQ figures and baselines
ISHR on UN80 mandate review modalities: working group timeline
Norway, statement on the working group progress report: ongoing process





Comments