Reading Time: 7 minutes

Eleven years into the Sendai Framework for Disaster Risk Reduction 2015–2030, disability inclusion remains the most-cited and least-implemented commitment in the entire disaster risk reduction (DRR) architecture. The 2025 mid-term review at the UN General Assembly’s High-Level Meeting in February named the gap in plain language, and the disaster record of the intervening twenty-four months — Türkiye-Syria, three Pacific cyclone seasons, the third year of Ukraine displacement — supplied the operational detail. This report synthesises what the data says about disability inclusion in disaster preparedness in 2026, scored against the floor set by the Sendai Framework, Article 11 of the CRPD, and the 2019 IASC Guidelines.

Persons with disabilities die in disasters at up to four times the rate of the general population, yet fewer than 11% of humanitarian financing flows in 2024 were tagged as disability-inclusive against a global disability prevalence of approximately 15% — the WHO 2024 estimate of 1.3 billion people, or one in six. The three-figure gap between the population, the funding, and the mortality outcome is the survey’s headline finding, and it is the metric the Sendai mid-point asked governments to close before 2030.

Overview of disaster-disability-inclusion compliance

Disability inclusion in disaster preparedness rests on three load-bearing instruments. The Sendai Framework, adopted at the Third UN World Conference on DRR in March 2015, is the only universal DRR instrument and the only one with an explicit, repeated disability-inclusion mandate; persons with disabilities are named in Priorities for Action 1, 2, 3 and 4, and Targets E, F and G are formally disaggregated by disability in the Sendai Monitoring Framework that UNDRR operates. Article 11 of the UN Convention on the Rights of Persons with Disabilities (CRPD), in force since 2008, requires States Parties to take all necessary measures to ensure the protection and safety of persons with disabilities in situations of risk. The Inter-Agency Standing Committee (IASC) Guidelines on the Inclusion of Persons with Disabilities in Humanitarian Action, adopted in November 2019, set the actual sector-by-sector standards humanitarian actors are expected to meet.

Against that floor, five failure modes recur across the 2024–26 evidence base:

  • The early-warning multimodality gap — operational, fully multimodal Common Alerting Protocol delivery (SMS plus accessible push, audio siren, visual strobe, and sign-language video) exists in fewer than 30 of the roughly 130 CAP-implementing countries.
  • The institutional-evacuation gap — residential institutions for older people and persons with intellectual or psychosocial disabilities routinely have no functioning evacuation plan, concentrating the people least able to evacuate into a single building.
  • The accessible-shelter and accessible-evacuation-centre gap — door widths, ramp gradients, and accessible WASH facilities in temporary settlements fail IASC Sphere-aligned standards in most post-event audits.
  • The displacement-disaggregation gap — registration mechanisms in receiving states consistently fail to capture disability on intake, so the displaced disabled population is invisible in the operational data.
  • The funding-marker gap — only about 11% of humanitarian financing is tagged disability-inclusive, and national DRR budgets specifically tagged for accessibility sit in the low single digits even in high-capacity reporting states.

The treaty floor is in place; the implementation gap is one of national-budget allocation, of who sits at the planning table, and of whether organisations of persons with disabilities (OPDs) are coordinated with as principals or briefed as stakeholders.

Breakdown by domain

The 2024–26 evidence sorts into four operational domains: early-warning systems, institutional residents and evacuation, recent case studies, and climate-adaptation finance.

Early-warning systems: multimodal CAP adoption

Sendai Target G — “substantially increase the availability of and access to multi-hazard early warning systems” — is the target with the clearest accessibility test, and the one where the gap is most measurable. The Common Alerting Protocol (CAP), an ITU-T standard (X.1303) that allows a single alert message to be issued in multiple modalities (text, audio, visual, sign-language video) simultaneously, is the field’s accepted technical answer.

The WMO–ITU CAP Adoption Tracker, updated through late 2025, listed CAP implementation in approximately 130 countries. Operational, fully multimodal CAP — the same alert delivered as SMS, accessible push notification, audio siren, visual strobe, and sign-language video on broadcast — is concentrated in fewer than 30. The remainder operate CAP only partially: SMS-based mobile alerting that excludes deaf users without visual confirmation; audio sirens that blind users can hear but that convey no instruction; visual-only alerts that deaf-blind users miss entirely. The UN Secretary-General’s Early Warnings for All initiative, launched in March 2022 with a target of full global coverage by end-2027, has made multimodal accessibility one of its formal workstreams; the 2025 mid-initiative report notes that of the 30 priority countries identified for initial scale-up, fewer than half had a national plan with explicit accessibility provisions for persons with disabilities as of mid-2025.

Institutional residents and evacuation

The IASC Guidelines specify stair-evacuation chairs, ramp-equipped shelters, and accessible WASH as the operational floor for sheltering and evacuating persons with disabilities. Post-event audits across the 2023–25 reporting cycle find those provisions absent from the public-building inventories of most affected jurisdictions. Where deinstitutionalisation reforms have progressed prior to a disaster, the risk profile of the disabled population is measurably lower, because community-based services prove more resilient than institutional ones. The recurring policy implication, named in the European Disability Forum’s 2024 joint statement with Inclusion Europe, is that deinstitutionalisation is disaster risk reduction: a residential institution in the path of a flood, a wildfire, or a front line concentrates the people least able to evacuate themselves into a single building with no plan for them.

Türkiye–Syria, 6 February 2023

The 7.8-magnitude earthquake that struck the Türkiye–Syria border in the early hours of 6 February 2023, followed by a 7.5 aftershock the same day, killed more than 59,000 people and displaced over 3 million. Post-event reviews — conducted across 2023 and 2024 by the Disability-Inclusive Disaster Response Coalition, the International Disability Alliance, Human Rights Watch, and the Turkish Confederation of the Disabled (Türkiye Sakatlar Konfederasyonu) — produced the most-detailed operational record the field has had since the 2011 Tōhoku and 2015 Nepal events.

The recurring findings: residents of residential institutions in the affected provinces had no functioning evacuation plan; deaf residents missed the morning’s AFAD (Disaster and Emergency Management Authority) SMS-based warnings entirely because the alerts were audio-only on most networks; wheelchair users were excluded from upper-floor evacuations in collapsed-but-standing buildings because the stair-evacuation chairs the IASC Guidelines specify were absent from almost every public building inventoried. Temporary container settlements deployed across the first six months were largely inaccessible. Reconstruction tenders issued by the Turkish Housing Development Administration (TOKİ) during 2024 began to incorporate accessibility specifications, but the coalition’s 2025 follow-up review found that as-built compliance lagged the tender language by a substantial margin. The Syrian side of the border, where the response was constrained by sanctions, divided territorial control, and the prior decade of conflict-driven institutional collapse, produced almost no operational disability data — itself a finding the coalition’s reports underline.

Pacific cyclone seasons 2023–24 and 2024–25

The Pacific cyclone seasons of 2023–24 and 2024–25 were the field’s clearest demonstration of what locally-led, OPD-coordinated disaster response looks like in practice. Cyclone Lola, which struck Vanuatu as a Category 5 system on 24–25 October 2023 — the earliest-recorded Category 5 in the Southern Hemisphere season — was followed by Cyclones Judy and Kevin (March 2023, retrospectively), Mal (November 2023), and a sequence of late-2024 systems including Kong-Rey’s edge effects.

The Vanuatu Disability Promotion and Advocacy Association (VDPA), working with the National Disaster Management Office and the Pacific Disability Forum (PDF), ran a disability-inclusive response model that other Pacific NDMOs have begun to copy. The model has three operational components: a pre-positioned register of persons with disabilities at province and area-council level, maintained with consent and used only by trained NDMO and VDPA personnel; community-level disability focal points trained on multimodal alert dissemination, who carry the early warning into the last mile when SMS and radio do not reach; and accessible-evacuation-centre audits conducted jointly with the Public Works Department in the lull between cyclone seasons. The model is imperfect — the registers are incomplete in remote outer-island communities, and the audit programme has lagged the schoolbuilding programme — but it is the closest functioning example of Sendai Priorities 2 and 4 in a small-island context. The PDF’s 2024 regional review noted that Fiji, Tonga and Solomon Islands have begun adapting elements of the Vanuatu model, with mixed progress on funding the focal-point networks at scale.

Ukraine: displacement and the institutionalisation multiplier

Pre-invasion baselines from the State Statistics Service estimated approximately 2.7 million persons with registered disability status in Ukraine, with the actual prevalence figure (using the WHO 15% baseline) closer to 6 million. The UN Refugee Agency’s registered displaced-population data, updated through 2025, shows roughly 6.8 million Ukrainian refugees recorded across Europe and a further estimated 3.7 million internally displaced; disability prevalence within those populations is consistently undercounted because the registration mechanisms in receiving states do not capture it on intake.

The operational record from 2022 through 2025 has been documented in detail by Human Rights Watch, the European Disability Forum, and the National Assembly of People with Disabilities of Ukraine (NAPD). Three findings recur. First, evacuation of residential institutions for older people and persons with intellectual or psychosocial disabilities in the path of advancing front lines was, in the first months, conducted via paper “evacuation lists” with no clear locus of responsibility between the Ministry of Social Policy and regional administrations. Second, accessible-shelter provision in Western Ukraine and in receiving states was a binding constraint throughout 2022–23 and remained partial in 2024. Third, where deinstitutionalisation reforms had progressed prior to 2022 — most notably in some western oblasts — the disaster-risk profile of the disabled population was measurably lower.

Climate-adaptation finance and the Loss and Damage Fund

The Loss and Damage Fund (formally the Fund for responding to Loss and Damage), agreed at COP27 in November 2022, operationalised at COP28 in December 2023, and with its board’s first formal disbursement decisions taken across 2024 and 2025, is the newest piece of the architecture. Disability inclusion was raised in the fund’s governance design — by the International Disability Alliance and a coalition of OPDs from climate-vulnerable states — but the founding instruments do not name disability as a cross-cutting commitment, and the initial project pipeline approved at the board’s 2025 meetings did not contain explicit disability-inclusion line items beyond generic “vulnerable groups” language. The advocacy ask going into the fund’s 2026 replenishment cycle is for the kind of named, budgeted disability-inclusion provision that the Sendai monitoring framework already nominally requires.

Quantitative insights

Read together, the 2024–26 monitoring data produces a consistent set of percentages:

  • Up to 4× — the disaster mortality multiplier for persons with disabilities, the figure most commonly cited in the Sendai mid-term review and the UN ESCAP 2024 Asia-Pacific Disaster Report, drawing on post-event mortality reviews from the 2011 Tōhoku earthquake forward.
  • 11% of humanitarian financing tagged disability-inclusive in 2024, the share reported in Development Initiatives’ Global Humanitarian Assistance Report 2024, against a global disability prevalence baseline of 15%.
  • 15% of the global population — approximately 1.3 billion people, or 1 in 6 — live with significant disability per the WHO 2024 baseline, the figure against which any operational disability-inclusion benchmark should be read.
  • Fewer than 40 reporting states had submitted any disability-disaggregated data on local DRR strategies as of the 2023 Sendai reporting cycle, despite Targets E, F and G being formally disaggregable by disability.
  • Consultation with persons with disabilities in national DRR strategy development was self-reported by 70-plus states but verifiable in fewer than half of those cases, per UNDRR’s mid-term review report.
  • Roughly 130 countries have implemented CAP; fewer than 30 operate it as fully multimodal, per the WMO–ITU CAP Adoption Tracker (late 2025).
  • Of the 30 Early-Warnings-for-All priority countries, fewer than half had a national plan with explicit disability-accessibility provisions as of the 2025 mid-initiative report.
  • National DRR budgets tagged for accessibility sit in the low single digits even in high-capacity reporting states.

Taken together: a population share of 15%, a mortality multiplier of up to 4×, a funding tag of 11%, and a national-budget allocation in the low single digits. The numbers describe a single structural shape — a population disproportionately exposed to disaster mortality, disproportionately under-resourced in disaster preparedness, and disproportionately invisible in the disaster data.

What good policy looks like in 2026

The 2025 mid-term review’s Political Declaration of 19 May 2023 reaffirmed the disability-inclusion language and added two specific new lines: a call for Common Alerting Protocol adoption with multimodal alerting, and an explicit reference to the IASC Guidelines as the operational floor. The countries doing this well share five features, not one: (1) a national DRR strategy that names disability inclusion with measurable indicators, not aspirational language; (2) OPDs at the DRR coordination table from strategy design through after-action review; (3) multimodal CAP-compliant early warning with audited accessibility in all four modalities; (4) accessible-shelter and accessible-evacuation-centre standards integrated into the building code and audited between events; and (5) a deinstitutionalisation pathway treated as part of the DRR portfolio, not a separate social-policy track.

Three country examples show what this looks like operationally. Bangladesh’s Cyclone Preparedness Programme (CPP), jointly run by the Government of Bangladesh and the Bangladesh Red Crescent Society since 1973, has incorporated disability-inclusion training in its 76,000-strong volunteer network since 2018 and works with the National Forum of Organisations Working with the Disabled (NFOWD) on multimodal warning dissemination in coastal districts. The Philippines’ Office of Civil Defense (OCD), under Republic Act 10121, has formalised inclusion of disabled people’s organisations in its DRR councils at national and regional level; implementation at municipal level remains uneven. Vanuatu’s VDPA-NDMO model is described above. For deeper background read the CRPD glossary entry, the national regulations index, and the wider 2026 reporting record.

Call to action for disaster planners and funders

The treaty floor names what disability-inclusive disaster preparedness requires; what the survey data shows is that the gap is one of allocation and inclusion at the planning table. Concrete next steps for 2026:

  • Move to fully multimodal CAP delivery. Audit the four delivery channels (SMS plus accessible push, audio siren, visual strobe, sign-language video on broadcast) and close whichever modality is missing from the national alert chain.
  • Seat OPDs at the DRR coordination table as principals. Inclusion in strategy design, contingency planning, and after-action review — not as stakeholders briefed after the fact.
  • Disaggregate the Sendai Monitoring data by disability. Targets E, F and G are formally disaggregable; submitting the disaggregated data is what makes the framework operational.
  • Treat deinstitutionalisation as DRR. Cross-walk the social-policy deinstitutionalisation pathway with the DRR portfolio; institutional concentrations are a measurable disaster-risk multiplier.
  • Tag disability inclusion in the humanitarian-financing markers. Move the 11% disability-inclusive financing share toward the 15% prevalence baseline, and require a named disability-inclusion line item in Loss and Damage Fund project approvals from the 2026 replenishment forward.
  • Audit accessible-shelter and accessible-evacuation-centre compliance between events. Integrate the IASC Sphere-aligned standards into the building code, and run joint audits with OPDs in the lull periods, on the Vanuatu model.
  • Capture disability on intake for displaced populations. Build a Washington Group Short Set question into refugee and IDP registration so the displaced disabled population becomes visible to the response.

Conclusion

The Sendai Framework, Article 11 of the CRPD, and the 2019 IASC Guidelines collectively say what disability-inclusive disaster preparedness looks like in enough operational detail that no country in 2026 can claim it does not know. The 2025 mid-term review, the Türkiye-Syria post-event record, the Pacific cyclone seasons, and the Ukraine displacement data show that the gap is one of national-budget allocation, of who sits at the planning table, and of whether OPDs are coordinated with as principals or briefed as stakeholders. Closing it before 2030 is what the Sendai mid-point asked for. Whether the next five years deliver it is, again, a national-budget decision.