A handoff preceded the promised repair
Did FEMA equip local partners for disability-access responsibilities before changing how it delivers disaster assistance? A chronology assembled by What’s the Scoop With Broach from a new federal audit and its attached agency response shows the opposite sequence for one important training resource. FEMA suspended its partner course in March 2025, shifted its recovery-center approach toward state and local operations in May, and stopped sending assistance teams door to door in June. The partner course was still unavailable in August 2026.
Our examination also separates three future commitments that could otherwise be mistaken for one near-term fix. A November 2026 course discussed in the audit concerns FEMA's own newly combined Civil Rights cadre. The agency's September 16 written response sets September 30, 2027 as the estimated completion date for restoring and strengthening training support for state and local partners. Its plan to solicit and incorporate disability-related feedback on the assistance application process carries a May 31, 2028 completion estimate.
That sequence establishes an oversight gap between changed service delivery and planned corrective work. It does not prove that every local partner lacked disability expertise, that all training stopped, or that a particular applicant was denied aid because the course was removed. FEMA describes continuing guidance, specialist assistance and accessibility equipment, and it agreed with all three GAO recommendations.
What changed, and when
The suspended course, Including People with Disabilities in Disaster Operations, had been freely available through FEMA's Emergency Management Institute to partners including state and local emergency managers. Officials told GAO it was removed for an internal review of compliance with executive orders. GAO's August 2026 status places its unavailability about 17 months after the March 2025 suspension. This is the status recorded by the audit, not a claim that this publication tested every FEMA course portal today.
FEMA's May 2025 memorandum said the agency would focus on supporting state- or locally operated recovery centers, reducing the need to establish FEMA centers and optimizing locally led recovery. FEMA told auditors it would continue supplying Communication and Accessibility Kits for its work when state and local governments could not supply their own. Those kits can include sign-language interpretation technology and braille information.
The June 2025 change affected door-to-door visits by Disaster Survivor Assistance teams. FEMA said fewer than 2% of all Individual Assistance applicants applied through those teams between January 2024 and June 2025. Officials said teams could instead serve multiple survivors from central community locations such as parks. That is a change in deployment, not evidence that every assistance team was abolished.
| Date | Documented action or status |
|---|---|
| March 2025 | Partner disability course suspended for executive-order review |
| May 2025 | Recovery-center support shifted toward state and local operations |
| June 2025 | Door-to-door assistance visits discontinued |
| August 2026 | Partner training remained unavailable |
| November 2026 | Planned first course for FEMA's combined Civil Rights cadre |
| September 30, 2027 | Estimated completion for partner-training support and regional-role standardization |
| May 31, 2028 | Estimated completion for application-feedback corrective work |
A small overall channel can matter differently to different applicants
GAO's application-method chart gives useful context for FEMA's low-use explanation, but it covers a different period. For January 2022 through December 2025, 23% of applications in the chart's reported-disability group used a call center or recovery center, compared with 16% in the no-reported-disability group. The seven-percentage-point difference shows greater use of those staff-assisted channels within the displayed disability group.
The chart reports door-to-door teams at 3% for the reported-disability group and 2% for the other group. Internet or mobile applications remained the majority in both: 75% and 82%, respectively. Rounded percentages can exceed 100 when added; they should not be converted into exact applicant counts. The chart shows approximately 1.3 million and 5.6 million applications in its two groups, rather than the complete population discussed elsewhere in the audit.
This comparison does not contradict FEMA's fewer-than-2% figure: the agency's explanation uses a shorter January 2024–June 2025 window, while the chart spans four calendar years and includes only a partial final year of home visits. Nor does the chart prove how applicants would behave after the change. It does establish that an overall utilization rate can conceal differences relevant to an accessibility review.
GAO's interview evidence explains why a channel with relatively few users can still matter. Stakeholders described inaccessible transportation, disrupted phone and internet service, medical reasons for remaining at home, and difficulty using crowded recovery centers. These were accounts collected by GAO, not interviews conducted by Billy Roach. They illustrate possible barriers; the selected interviews do not measure their national prevalence.
The written response schedules different repairs on different clocks
In its response to the first recommendation, DHS said FEMA would review its materials, technical-assistance resources and partner-developed courses, identify updates and gaps, and decide how best to make training available. It said the agency would prioritize jurisdictions without comparable training. The September 30, 2027 estimate is a completion target for that work; the letter does not say that no interim assistance can occur before then.
The second recommendation concerns standardizing regional disability specialists' responsibility to coordinate with relevant partners before disasters. DHS gave that work the same September 30, 2027 estimate while preserving flexibility for differences among regions. The audit found three of ten regional specialist positions vacant as of May 2026 and described uneven partner awareness of the role. Those are dated observations, not a verified October staffing count.
The third response sets a longer sequence: seek public comment on the Individual Assistance Program and Policy Guide in fall 2027, consider feedback including disability-related application barriers, and publish an updated guide by spring 2028, with an estimated completion date of May 31. That is a plan to evaluate and incorporate feedback as appropriate—not a promise that every obstacle identified by GAO will be eliminated.
These dates must be kept separate from the November 2026 internal course. FEMA merged its deployable Disability Integration and Civil Rights cadres in 2025, and officials described that course as training for the combined role. Treating it as restoration of the public partner course would give readers an earlier resolution date than the written response actually provides.
Improvements already made do not resolve the chronology
The audit credits FEMA with changes that help applicants. Officials described a more visual online application, clearer progress indicators and multiple plain-language techniques. FEMA also revised determination letters and simplified appeals. Officials said merging cadres increased the number of staff available to serve people with disabilities, although one former disability-cadre member told GAO the new role had become more administrative.
At the same time, GAO found that application definitions for assistive devices and medically required equipment appeared only after an applicant selected Yes. It reported that FEMA's March 2025 focus groups for determination letters did not specifically include people with disabilities. These are concrete reasons GAO recommended obtaining feedback directly from people with different disabilities, rather than assuming general usability work captured their needs.
This investigation cross-references the audit's training, delivery-channel and staffing sections with figure 6 and the signed September 16 DHS response in appendix II. Our original contribution is the reconciled sequence and the distinction among the 2026, 2027 and 2028 commitments. GAO examined selected disasters and stakeholders in Hawaii, Kentucky and Louisiana; those interviews are not a representative national survey. The records establish changes and planned deadlines, but do not quantify assistance lost because of them. We have not claimed firsthand disaster visits, new agency contact or an October determination that the planned work is complete.
Sources and further reading
GAO-26-108057, September 30, 2026: full audit and DHS response ↗
Original report PDF; DHS September 16 response on printed pages 36–39 ↗
FEMA outreach photograph, caption and public-domain rights ↗
