Separations were large relative to the workforce
GAO reported that FEMA employed an average of about 25,134 people in fiscal 2025 and experienced more than 4,300 separations. Dividing the minimum reported separations by the average workforce produces 17.1%—roughly one separation for every six average positions.
That ratio is a scale comparison, not a formal turnover rate: the denominator is an annual average, the numerator is described as more than 4,300, and FEMA could hire people during the same period. GAO separately reported that separations rose 55% from fiscal 2024.
| Indicator | GAO finding |
|---|---|
| Average FY 2025 workforce | About 25,134 |
| FY 2025 separations | More than 4,300 |
| Separations as share of average workforce | At least 17.1% |
| Increase from FY 2024 | 55% |
| FEMA recommendations | 3; all open on GAO page reviewed Sept. 23 |
The reductions lacked the planning process GAO expected
GAO found FEMA did not base its 2025 and 2026 workforce reductions, proposed future actions or policy changes on an analysis of current staffing and forecast mission needs. FEMA had also rescinded its strategic plan in 2025.
Agency officials told GAO the departures cost FEMA institutional knowledge and experienced personnel. GAO interviewed officials at headquarters and all 10 regions, analyzed workforce data and compared planning with Office of Personnel Management practices.
DHS agreed, but the corrective work remained open
GAO recommended a strategic plan, an agency-wide workforce analysis and an annual staffing plan informed by that process. DHS concurred with all three. The public recommendation table still lists all three as open, even after the Senate confirmed a FEMA administrator in August 2026.
Concurrence is not completion. DHS said some work would wait for strategic-plan outlines; GAO continued to warn that delay could leave FEMA underprepared and under-resourced.
What the evidence does—and does not—show
The original contribution is to put the departure count on the scale of FEMA’s average workforce and connect it to the still-open planning controls. The finding is about readiness governance, not a claim that a specific disaster response failed.
GAO did not say every person who left held a critical response role, and the average headcount includes different employment categories. Mission impact depends on occupation, experience, deployment availability and replacement hiring—details the 17.1% comparison cannot supply.
Sources and further reading
GAO-26-108427 findings, agency response and recommendation status ↗