A Mandate Without a Lever: What the Measured Record Says About the UN Pact for the Future's Space Provisions

Key Findings

• Binding treaty, measured: 9,589 of 10,244 eligible payloads launched 2015–2023 were registered under the 1975 Registration Convention — 93.6 per cent. The pooled tail of states launching fewer than 40 payloads sits at 56.3 per cent across 531 payloads. Both are our own counts against the GCAT catalogue on a 16 August 2026 snapshot.
• Voluntary guideline, measured: active upper-stage disposal rose from 2.8 per cent of orbital launches in 1995–99 (11/394) to 61.2 per cent in the partial 2025–26 window (318/520). Of those 318 disposed launches, 263 (82.7 per cent) were US. The rest of the world sits at 22.8–25.9 per cent and has barely moved since 2020.
• Twenty-fold spread under one instrument, 2020s: United States 82.3 per cent, France 71.1 per cent, Japan 25.0 per cent (6/24), China 21.2 per cent, Russia 20.7 per cent, India 3.8 per cent (1/26).
• No machinery on either side: neither instrument has verification, a compliance committee, a review conference or a sanction. COPUOS’s 69th session (June 2026) failed to adopt its report by consensus, leaving no agreed 2027 agenda. At that same session its own Space Situational Awareness Expert Group failed consensus on its draft recommendations.
• The Pact’s one concrete space instruction failed inside two years: Action 56’s chapeau (September 2024) invited COPUOS to consult on holding UNISPACE IV in 2027. COPUOS declined to agree it, with Austrian funding on the table.

Sources and evidence strength for every figure above: see the Key Figures table at the end.

Executive Summary

This analysis assesses the space provisions of the Pact for the Future against the two instruments any new mandate would have to improve on: one binding, one not. The dominant force in this environment is a widening divergence. On one side sits a legal layer whose obligations go unenforced and unverified; on the other, a technological layer that changed by an order of magnitude inside a single guideline cycle. The direction of travel is hostile for the multilateral track and permissive for functional coordination outside it.

The Landscape

Context and Scope

The Pact for the Future was adopted in September 2024 as A/RES/79/1, with 56 Actions. The space commitment is Action 56, the last of them. The scope here is global and multilateral, with jurisdiction-level comparison across the United States, France, China, Russia, Japan, India, Italy, Spain and a pooled tail of small launching states. The venues in scope are COPUOS and its subcommittees, UNOOSA and the ITU, alongside the non-UN operative layer of TraCSS, Space-Track and bilateral arrangements. The instrument baseline runs 1961 to 2019. The measured behavioural windows are 1995–2026 for disposal and 2015–2023 for registration, and the institutional record runs through the 69th COPUOS session of June 2026. The focal question is whether the Pact constitutes a new mandate, or ratifies a shift already under way from treaty-making to technical coordination. Behind it sits a second: what measured behaviour says about which instrument type actually works.

One clarification about the jurisdictions named here, because the two measurements do not count the same states. On the launch side the catalogue attributes each launch to the state of the launching organisation, so Europe appears almost entirely as France. Arianespace is a French company, and it flies from the Centre Spatial Guyanais at Kourou, in French Guiana. Across the 2020s France records 45 orbital launches, Germany one, and no other European state any. Thirty-four of those 45 are Arianespace flights from Kourou, thirteen of them Vega and Vega C vehicles built at Colleferro in Italy. The remaining eleven are Starsem flights of a Russian vehicle from Vostochny and Baikonur, counted as French because Starsem is French-registered as well. France on this axis is Europe’s launch industry rather than French national activity. The absence of Germany, the United Kingdom, Italy and Spain reflects the absence of a national launch operator rather than absence from orbit. Registration is measured on a different axis, the state of registry, and there those same states appear in their own right: the United Kingdom is the third largest in the dataset after the United States and China, with 700 payloads requiring registration against France’s 49.

The Macro-Environment at a Glance

The Operating Environment

Start with the binding instrument, because it is the one that can be measured against its own text. The Registration Convention obliges launching states to furnish the UN with particulars of every object they place in orbit. Counting the payloads launched between 2015 and 2023 that the catalogue judges to require registration, 93.6 per cent were registered. No source in the literature reports that figure. It is our own count against Jonathan McDowell’s General Catalog of Artificial Space Objects , on a single 16 August 2026 snapshot and under one eligibility judgement. UNOOSA’s own headline of roughly 85 per cent points the same way. But it covers all object classes across all time, a different population, so the two corroborate each other in direction only and never as a trend. Beneath the aggregate, the distribution is what matters: Uruguay at 100 per cent on 44 payloads, the United States at 98.7 per cent on 7,274, China at 79.7 per cent on 867, then Italy at 58.0 per cent on 50 and Spain at 52.5 per cent on 40, with the pooled small-state tail worst at 56.3 per cent.

That shortfall is a drafting outcome rather than an aberration of practice. Article IV(1) requires information “as soon as practicable” and sets no deadline, so late filing is formally compliant. Article II(3) leaves the contents and conditions of each national registry to the state’s own discretion, so partial filing is compliant too. The end-of-life notification in Article IV(3) carries the treaty’s weakest standard, “to the greatest extent feasible and as soon as practicable”, which makes any decay in disposal reporting legally costless. Meanwhile the transparency principle the current era believes it is inventing was written into the treaty half a century ago: Article III(2) states that there shall be full and open access to the information in the Register. The deficit is in deposits, not in permissions. There is no review conference, no implementation meeting, no compliance committee and no dispute resolution attached to any of it.

The regime also excludes its own data sources. Registration is valid only through Permanent Missions or IGO headquarters , so agencies, universities and companies — the actors that generate the information — cannot validly file it. The obligation is assigned to the party furthest from the data. The consequence is visible in who goes missing: unregistered payloads are overwhelmingly cubesats, university missions and small commercial units, with 176 Chinese, 91 US, 21 Italian and 11 French among them. Deliberate concealment exists, but it is rare and identifiable: a 2018 academic count found 17 apparently intentional omissions against 71 accidental . That same count found a systematic military blind spot: as of 2018, no State Party had ever described a registered object as having a military function. The dominant pattern is capacity-limited non-performance. Italy at 58.0 per cent and Spain at 52.5 per cent are not defying anything. The pattern is old, too. CNES reported in 2005 that 25 per cent of more than 250 Ariane-launched satellites were unregistered.

Now the voluntary instrument, and the result that inverts the expectation. Active disposal of upper stages was recommended by the 2007 COPUOS debris-mitigation guidelines and reinforced by the 21 Long-Term Sustainability Guidelines adopted in 2019 . It rose from 2.8 per cent of orbital launches in 1995–99 to 61.2 per cent in the partial 2025–26 period, on the same catalogue and the same snapshot. That headline may not stand alone. Of the 318 disposed launches in the recent window, 263 were United States launches, 82.7 per cent of the total. Excluding the US, the rest of the world sits between 22.8 and 25.9 per cent, essentially flat since 2020. Unlike the registration figure, this one has no independent numerical check anywhere in the corpus. Byers and Boley reach the same caution by a different method, warning that the improvement appears driven mostly by SpaceX’s own practices and may be diluting averaged metrics. They corroborate the caveat rather than the value. The rule-makers themselves were more pessimistic still: the Inter-Agency Space Debris Coordination Committee concluded in 2017 that adherence to its own guideline set was insufficient with no apparent trend towards better implementation.

The instrument is the same everywhere; the behaviour is not, and that is what isolates the driver. Across the 2020s the United States sits at 82.3 per cent and France at 71.1 per cent, with everyone else clustered at roughly a quarter or below: Japan 25.0 (6/24), China 21.2, Russia 20.7, India 3.8 per cent (1/26) — the smallest denominators in the set, where a single launch moves the rate by several points. The two leaders are the two jurisdictions with the clearest domestic transposition, the FCC’s five-year rule of 2022 and French licensing under the 2008 space operations law. That reading holds better for the United States than for France. Eleven of France’s thirty-two disposals are Starsem flights of a Russian vehicle from Vostochny and Baikonur, where French licensing is not what decides the upper stage’s fate, and Arianespace’s own launches from Kourou sit at 61.8 per cent rather than 71.1. India’s Space Activities Bill of 2017 remained pending. One figure in that series must be read carefully. New Zealand’s low rate reflects Rocket Lab Electron flights of a US-owned company, since the grouping is by launching organisation’s state, and it is not evidence about New Zealand policy.

That spread also points to stringency being the wrong lever. NASA’s Orbital Debris Program Office models a 200-year horizon. With no adherence to the 25-year disposal rule, debris grows by about 330 per cent; at roughly 90 per cent adherence, growth holds near 110 per cent. Tightening the rule to five years while holding adherence at that level improves the outcome only marginally, to about 100 per cent. These are model outputs from a single source rather than measurements, and the precision should not be read as earned. The qualitative result is robust, and it is what carries the argument. J.-C. Liou, chief scientist of NASA’s Orbital Debris Program Office, put it plainly : good policies and practices are in place, and the global record on implementing them is poor.

Against that evidentiary background, the Pact reads differently than its billing suggests. Action 56 commits states to “discuss the establishment of new frameworks for space traffic, space debris and space resources” through COPUOS. The operative verb is discuss. Its first limb reaffirms the importance of the widest possible adherence to and full compliance with the 1967 Outer Space Treaty. That instrument has no verification methodology in the open literature , no review conference and no implementation machinery. So the Pact’s one quasi-obligatory gesture on space is a demand for full compliance with something nothing can check. Its second limb invites private-sector and civil-society engagement “where appropriate and applicable”, narrower than the zero draft’s language. Action 27 preserves the commitment to prevent an arms race in outer space , anchored to the Outer Space Treaty, with the earlier General Assembly resolution on the subject absent from the adopted text. The one concrete instruction is in Action 56’s chapeau, encouraging COPUOS to consult on holding a fourth UNISPACE conference in 2027. COPUOS could not agree to it, even with Austria offering to fund it. Stimson tracks the Pact most systematically, across 452 sub-action indicators drawn from 40 of the 56 Actions. It includes no space Action among the ten flagship commitments it proposes for the 2028 review.

The venue receiving that mandate is deteriorating faster than the mandate is arriving. The 69th session failed to adopt its report by consensus, leaving no agreed agenda for COPUOS or its subcommittees in 2027. Membership has reached 110 states plus 62 permanent observers with four applications pending, and the Secure World Foundation assesses that growth as making consensus harder and diluting agreed substance . The failure mode is now vocabulary: a United States objection to the phrase “developing countries” blocked the session despite multiple alternative formulations. The sharpest test is whether the technical layer escapes any of this, and it does not. At that same session the Space Situational Awareness Expert Group could not reach consensus on its own draft recommendations, and the Long-Term Sustainability Working Group deferred its work plan to February 2027.

Scarcity is doing as much of this work as diplomacy. The UN liquidity crisis cut the 69th session from eight days to seven, capped statements at three to four minutes and pushed substance into informal, English-only consultations. The 68th session had already run three-minute statements with no technical presentations. Aarti Holla-Maini, Director of UNOOSA, told the session that resource cuts have degraded the office’s ability to perform treaty-mandated functions, specifically including updating and maintaining registration information for space objects. Reported shortfalls run as high as 30 to 40 per cent in some UN bodies, and official development assistance fell 23.1 per cent between 2024 and 2025. Regulatory capacity, meanwhile, is being capitalised elsewhere. Europe is coupling €131bn for defence and space in its multiannual framework to an explicit governance-reform agenda , and the EU Space Act is argued to make hard-law options previously considered untenable more available to governments . A new mandate would land on a secretariat already stating it cannot maintain the register an existing treaty obliges it to keep.

The participation cost of all this is rarely priced. Bahrain stated on the record that governance arrangements for space situational awareness and traffic remain largely non-binding, fragmented and dependent on voluntary cooperation , and separately that it operates no indigenous sensors and maintains no national catalogue. For most members, information sharing means receiving. SIPRI notes that many states lack the personnel, expertise, time or funds to engage across proliferating fora. A Stimson retreat found smaller missions unable to track multiple negotiation tracks, with think tanks filling the analytical gap. The Secure World Foundation distributed more than 200 copies of its COPUOS briefing book in four languages at one Legal Subcommittee session, citing a knowledge deficit. English-only informals fall hardest on the newest members whose accession drives the membership growth. The data contests the equity framing as well: Starlink users are concentrated in the wealthiest countries , and the Philippines welcomed the Pact’s space language for disaster risk reduction rather than for orbital safety. The entities operating most of the hardware have no standing at all. Commercial operators entered the 2024 sustainability workshop only as state-nominated panellists whose views are explicitly not state positions.

The technological layer explains why the multilateral clock is losing. Satellites grew from 852 in 2004 to more than 9,000 in 2023 . SpaceX reported some 50,000 collision-avoidance manoeuvres in six months of 2024 against no internationally agreed rules of the road, and intends 11,000 further satellites with filings for 30,000 more. A single operator’s filings therefore exceed the roughly 6,000 objects the UN register had accumulated in its first five decades, as of April 2016. The counter-model is already operating. TraCSS went from nine operators and about 1,000 satellites at beta in September 2024 to 68 pilot users, more than 11,290 satellites and nine National Government Accounts including Brazil and Egypt by July 2026. That is a de facto international arrangement assembled by administrative onboarding rather than ratification, in roughly 21 months. The Artemis Accords grew from 29 signatories to 55 by May 2025 , in about two years, on the same accession logic. The one UN-family instrument with teeth shows where the difference lies: UNOOSA characterises the ITU Radio Regulations as a binding international treaty , and the ITU enforces deployment milestones of 10 per cent within two years, 50 per cent within five and completion within seven, because it allocates a scarce resource operators need. No allocation lever, no compliance.

Where Factors Converge

Three reinforcing loops dominate, and all three operate inside the UN system. The first is a spiral of consensus and scarcity. Membership growth raises the cost of consensus. Budget scarcity compresses the deliberative practice that made consensus achievable. And compressed English-only informals exclude the newest members, weakening the legitimacy of whatever consensus survives. The 69th session’s failure to agree even a 2027 agenda is that loop’s observable output. The reform menu on the table, combining agenda items and shortening sessions, pointedly excludes the decision rule that drives it.

The second is frontier degradation. Falling launch costs admit an actor class the 1975 regime was never designed around. Those actors are procedurally barred from filing their own registrations, and their states are disproportionately those with the thinnest administrative practice, which is why the pooled small-state tail sits at 56.3 per cent. Each new entrant lowers aggregate compliance without any state deciding to defect. The regime degrades as the actor base broadens, which is the direction of travel for the coming decade.

The third is functional substitution. Because UN-level instruments cannot bind, capability is built outside the UN. Once TraCSS, Space-Track and bilateral channels supply the operative service, the marginal value of a UN instrument falls further, and states onboard administratively instead of negotiating. The United States states the sequencing openly: build the system, then ask COPUOS to find a home for the conversation .

Two dampening effects run the other way, and both operate outside the UN as well. Soft law hardens downstream: the 2007 guidelines acquired binding force through ISO in 2010, through ECSS in 2015 where they became mandatory for all ESA projects, and through the FCC’s five-year rule in 2022. That partially answers the free-riding objection, but only where such transposition exists. The twenty-fold measured spread in disposal rates is what the gap between “where it exists” and “where it does not” looks like. Binding rules are also arriving by a different route, through large-market regulators with extraterritorial reach. Both propositions hold at once: binding rules are coming, and they are not coming from the UN.

Three analytical lenses sit behind this reading: regime structure, institutional rules and cross-jurisdictional comparison. They converge. But they read the same two datasets and were chosen for their fit to that evidence, so their agreement is coherence rather than independent replication.

The Outlook

What This Means

For COPUOS and UNOOSA, the change with the most leverage available is not a new normative text but a change to who may file. The rule assigns the registration duty to the party furthest from the data. Permitting or requiring operator-level submission would address the structural cause of chronic under-registration directly. The second priority is the measurement layer. No agreed metric for implementing the sustainability guidelines has existed since the gap was first flagged, and independent behavioural data is currently the only scorecard anyone has. The third is uncomfortable, because it is the reform the Committee will not consider. The binding constraint is the consensus decision rule rather than the drafting, and every reform preserving universality plus consensus forecloses enforcement by construction.

For states and national regulators, the operative lever is domestic transposition with a licensing consequence, on the pattern that ran from ISO through ECSS to the FCC. The two top performers on disposal are the two jurisdictions with the clearest domestic instruments, and nothing in the record suggests treaty status predicts anything comparable. The corresponding risk is regulatory competition rather than regulatory absence. The flexibility that bought consensus on the sustainability guidelines is a documented arbitrage channel. It shows on the registration axis too, in the 94 payloads whose filing state differs from the catalogue’s owner attribution. That argues for international floor-setting rather than harmonisation. For newly launching states the constraint is administrative capacity, not compliance will, so support should be toolkit-based rather than enforcement-based.

For operators, obligations will arrive from large-market regulators with extraterritorial reach, faster than any multilateral text. Accession-based technical services impose terms rather than confer rights: TraCSS is supplied as is, with no warranty and with the US Government immune from suit . That is a service, not a legal regime.

One serious objection deserves stating rather than dismissing. CSIS has argued it is unclear whether the continued absence of a mission-authorisation framework would actually impact US companies pursuing novel space activities , and the question generalises: an implementation gap is not self-evidently a harm. The answer the evidence supports is narrow but firm. The harm is not the gap in the text. It is that the measurement and participation base is eroding while the actor population grows, which is the most consequential force in this environment. Rules whose adherence cannot be observed cannot be improved, and adherence, not stringency, is what the debris modelling says matters.

What to Monitor

Four signals would indicate the environment is shifting. Politically, whether COPUOS agrees a 2027 agenda and restores a functioning report cycle at its next session, observable in the Committee’s own report and in the Secure World Foundation’s session summaries. Legally, whether UNOOSA’s Registration Project moves beyond toolkits, ST/SPACE/91 and ST/SPACE/95, toward a filing channel open to operators, and whether the Online Index returns to service so the register can be independently interrogated. Environmentally, whether the non-US disposal rate moves off its 22.8 to 25.9 per cent plateau in the next catalogue window. That is the cleanest available test of whether the norm is diffusing or the global average is simply riding US launch volume. Economically and legally at once, the passage and extraterritorial scope of the EU Space Act, together with national transposition elsewhere, since that is the route by which binding obligation is actually arriving. The Space Situational Awareness Expert Group’s 2028 report is the fixed point against which all four can be read.

Limitations

The behavioural windows differ by dataset, 2015–2023 for registration and 1995–2026 for disposal, and the 2025–26 disposal period is partial and in progress. All catalogue figures are counts on a single 16 August 2026 snapshot. Registration records for 2024 onward measure how slowly the register fills, not compliance, and are excluded here. Upper-stage disposal at launch, satellite end-of-life deorbit adherence and high-orbit compliance are three distinct measures and are never combined. The institutional evidence base is COPUOS-internal by construction, since almost all of it is UNOOSA-hosted, so claims about what any major power wants rest on weaker secondary footing. No think-tank source in the corpus analyses the Pact’s text in relation to outer space, so the link from the Pact to COPUOS practice rests chiefly on the German and Emirati non-paper citing Action 56 as authority . National transposition status is evidenced only for the United States and France; its absence elsewhere is unevidenced rather than established.

Key Figures — Sources & Evidence

“Single source” below means the figure rests on one dataset or one outlet — directional, not settled. Every catalogue figure is our own count against GCAT on the 16 August 2026 snapshot and has no independent numerical replication in the corpus; UNOOSA’s all-time headline of roughly 85 per cent registration covers a different population and corroborates direction only. The row marked as a conflict is the one most easily merged by mistake: upper-stage disposal at launch is not the same measure as satellite end-of-life deorbit adherence or as high-orbit compliance, and the three may never be compared as if commensurable.

Primary Sources & Research

United Nations Office for Outer Space Affairs (n.d.). Convention on Registration of Objects Launched into Outer Space. UNOOSA. https://www.unoosa.org/oosa/en/ourwork/spacelaw/treaties/registration-convention.html

United Nations General Assembly (2024). Pact for the Future (A/RES/79/1). United Nations. https://documents.un.org/doc/undoc/gen/n24/272/22/pdf/n2427222.pdf

United Nations Office for Outer Space Affairs (n.d.). United Nations Register of Objects Launched into Outer Space. UNOOSA. https://www.unoosa.org/oosa/en/spaceobjectregister/index.html

United Nations Office for Outer Space Affairs (n.d.). Registration of space objects — resources. UNOOSA. https://www.unoosa.org/oosa/en/spaceobjectregister/resources/index.html

United Nations Office for Outer Space Affairs (n.d.). The Registration Project. UNOOSA. https://www.unoosa.org/oosa/en/ourwork/spacelaw/capacitybuilding/registrationproject/index.html

United Nations Office for Outer Space Affairs (2019). COPUOS adopts guidelines for the long-term sustainability of outer space activities (UNIS/OS/518). UNOOSA. https://www.unoosa.org/oosa/en/informationfor/media/2019-unis-os-518.html

United Nations Committee on the Peaceful Uses of Outer Space (2026). Space situational awareness realities (A/AC.105/2026/CRP.11). UNOOSA. https://www.unoosa.org/res/oosadoc/data/documents/2026/aac_1052026crp/aac_1052026crp_11_0_html/AC105_2026_CRP11E.pdf

Germany and the United Arab Emirates (2025). Leveraging the Expertise of both Subcommittees to Address Urgent Questions related to Space Traffic — non-paper. UNOOSA. https://www.unoosa.org/documents/pdf/copuos/2025/GER_UAE_Non_Paper_Space_Traffic_23062025.pdf

United Nations Office for Outer Space Affairs (2022). Inter-Agency Meeting on Outer Space Activities — International Telecommunication Union. UNOOSA. https://www.unoosa.org/oosa/en/ourwork/un-space/iam/2022_itu.html

Trebaol, J.-Y. (CNES) (2005). French policy and practices for the registration of space objects. Fourth European Conference on Space Debris, ESA SP-587. https://conference.sdo.esoc.esa.int/proceedings/sdc4/paper/37/SDC4-paper37.pdf

Office of Space Commerce (2026). Traffic Coordination System for Space (TraCSS). U.S. Department of Commerce. https://space.commerce.gov/traffic-coordination-system-for-space-tracss/

McDowell, J. (2026). General Catalog of Artificial Space Objects. planet4589.org. https://planet4589.org/space/gcat

Secure World Foundation (2026). COPUOS at a turning point: Balancing a vital mandate with membership growth and global realities. Secure World Foundation. https://www.swfound.org/publications-and-reports/insight-jul2026-copuos-at-a-turning-point-balancing-a-vital-mandate-with-membership-growth-and-global-realities

Stimson Center (2026). Global Governance Innovation Report 2026: Advancing the Pact for the Future and a New Peace and Security Architecture. Stimson Center. https://www.stimson.org/2026/global-governance-innovation-report-2026-advancing-the-pact-for-the-future-and-a-new-peace-security-architecture/

Center for Strategic and International Studies (n.d.). Mission Authorization: Decoding the Space Policy Dilemma. CSIS. https://www.csis.org/analysis/mission-authorization-decoding-space-policy-dilemma

European Space Policy Institute (2025). Towards a Safe & Sustainable Space: Bridging Policy and Programmatic Ambitions. ESPI. https://www.espi.eu/news/towards-a-safe-sustainable-space-bridging-policy-and-programmatic-ambitions/

The Aerospace Corporation (2025). What will the EU Space Act’s new regulations mean for the industry?. Aerospace Kickstage. https://aerospace.org/kickstage/what-will-eu-space-acts-new-regulations-mean-industry

Observer Research Foundation (n.d.). Not-so-toothless norms for outer space. ORF. https://www.orfonline.org/expert-speak/not-so-toothless-norms-for-outer-space

arXiv:2512.20016 (2025). Verification of the Outer Space Treaty with Cosmic Protons. arXiv. http://arxiv.org/abs/2512.20016v4

arXiv:2512.22429 (2025). Chain Reactions in Space: Analyzing the Impact of Satellite Collisions and Debris Accumulation. arXiv. http://arxiv.org/abs/2512.22429v1