The Public Assistance indicators are published, adjusted every October, and widely misread as a pass mark. Individual Assistance has no equivalent number at all. This is what each one measures, and where the arithmetic stops and judgment starts.
Fiscal year 2026 figures, for incidents beginning on or after 1 October 2025. Verified against the Federal Register adjustment notices and the September 2025 Treasury release. Current as of 19 September 2026.
Not "was it bad." Not "did it cost a lot." Federal assistance under the Stafford Act is supplemental by statute. It cannot substitute for capability that already exists. So every factor FEMA weighs, on both the PA and IA side, is an attempt to measure damage against capacity rather than damage on its own. Once you see the factors that way they stop being a checklist and start being a single argument.
42 U.S.C. 5121 et seq. · 44 CFR 206.48These two programs measure different kinds of harm, and that difference is the whole reason one has a published dollar indicator and the other has six factors and no threshold.
Measures repair cost to publicly owned infrastructure and the cost of the emergency response. That damage is countable, uniform, and falls on a treasury.
Because the burden lands on government, population works as a rough stand-in for the tax base that would otherwise absorb it. Dividing damage by population produces a number that means roughly the same thing in Wyoming as in Texas.
Measures uninsured loss to homes and personal property. The same dollar figure means entirely different things depending on who absorbed it.
Two counties with identical flood damage diverge completely once you account for insurance penetration, homeownership rate, poverty, and age. Population tells you almost nothing about whether those households can recover on their own.
When FEMA rewrote the IA factors in 2019 it considered adding thresholds and said plainly why it would not:
FEMA will not be using a threshold because it would unnecessarily limit FEMA's ability to advise the President and would not allow FEMA to fully consider all factors that may be relevant for the unique circumstances of a disaster and its impact on the State.Final rule, 84 FR 10664, 21 March 2019
There is also a statutory reason no number can be decisive on either side. The Stafford Act contains a one-sentence provision that shapes the entire framework:
No geographic area shall be precluded from receiving assistance under this chapter solely by virtue of an arithmetic formula or sliding scale based on income or population.42 U.S.C. 5163, Limitation on use of sliding scales (Stafford Act § 320)
This is why FEMA calls $1.94 an indicator and never a threshold, and why it can recommend PA for a state that fell short of it. FEMA addressed this directly when it adopted the cost-to-capacity ratio, noting it does not deny assistance based solely on the formula. The result is one factor among many. Any teaching that presents these numbers as a pass/fail line is teaching it wrong.
Public Assistance is the only side of the house with a published dollar trigger. FEMA takes the estimated cost of federal and non-federal PA across the whole state, divides by statewide population, and compares the result to a per capita indicator that gets adjusted every 1 October for CPI-U. The same indicator applies to every state and territory regardless of size, because the Stafford Act bars FEMA from using a sliding scale based on population or income.
Two things trip people up. First, this is an indicator, not a cutoff: FEMA has recommended PA below it on localized impacts, recent multiple disasters, or mitigation investment, and has declined requests that cleared it. Second, a separate $1 million statewide minimum has been FEMA practice since 1999 and has never been adjusted for inflation. It only binds in the four small territories where per capita math lands under seven figures.
Localities qualify for designation on their own countywide per capita figure, which is why a county can carry a declaration that the state as a whole would not have reached. Virginia's independent cities are treated as county equivalents here.
Prefilled with Washington County, Virginia. Change the population for any locality.
The recurring Southwest Virginia pattern is worth walking end to end, because it is the case where the arithmetic and the outcome pull in opposite directions.
Step one, the statewide picture. Total PA damage across all four localities comes to $3.1 million. Virginia's statewide threshold is $16,744,902. The event reaches 18.5 percent of the indicator, which is statewide per capita damage of $0.36 against an indicator of $1.94. On the headline number, this is not close.
Step two, the local picture. Every affected locality clears its countywide indicator, most of them by an order of magnitude.
| Locality | Population | Threshold at $4.86 | PA damage | Per capita | Met |
|---|---|---|---|---|---|
| Buchanan County | 20,355 | $98,925 | $980,000 | $48.15 | yes |
| Tazewell County | 40,429 | $196,485 | $1,150,000 | $28.44 | yes |
| Dickenson County | 14,124 | $68,643 | $640,000 | $45.31 | yes |
| City of Norton | 3,646 | $17,720 | $330,000 | $90.51 | yes |
| Virginia statewide | 8,631,393 | $16,744,902 | $3,100,000 | $0.36 | no |
Step three, what that combination means. Norton at $90.51 per capita is carrying damage more than eighteen times its own indicator. That is a severe local event inside a state large enough to dilute it to nothing statewide. The countywide figures are what the localized impacts argument is built from, along with the 24-month disaster history and whatever the state has already spent from its own funds this budget cycle.
Step four, the IA side. Say 180 homes take damage, flood insurance penetration is low, and the PDA produces an estimated IHP cost of $2.4 million. Virginia's TTR is $845.9 billion, so the cost-to-capacity ratio is 2.4M ÷ 845,900 = 2.84, the bottom historical band, where 43 percent of requests were approved across 2008 to 2025.
This is the part worth being precise about, because it gets asked as though a per capita number exists. It does not. FEMA's own declaration criteria state plainly that there is no per capita indicator established as a factor for the consideration of Individual Assistance. The March 2019 final rule that rewrote 206.48(b) considered setting thresholds and deliberately declined. The preamble carries a section titled "Lack of Thresholds."
Instead there are six factors, two of which FEMA designates as principal for the Individuals and Households Program.
Total Taxable Resources first, then GDP by state where TTR is unavailable or stale, then per capita personal income by local area. Also weighs NGO and private sector activity and the cumulative effect of disasters over the prior 24 months.
206.48(b)(1)Cause of damage relative to peril coverage, concentration of damage, homes impacted and degree of damage, estimated cost of assistance, homeownership rate, and the share of affected households carrying insurance appropriate to the peril.
206.48(b)(2)Poverty status, existing assistance receipt, pre-disaster unemployment, age 65+ and 18-, disability, and limited English proficiency, each compared against national averages.
206.48(b)(3)Disruptions beyond 72 hours to life-saving and life-sustaining services, essential community services, and transportation or utilities.
206.48(b)(4)Missing, injured, or deceased, read as an indicator of community trauma and of likely crisis counseling and funeral assistance need.
206.48(b)(5)Survivors who lost work and do not qualify for regular state UI, typically self-employed, service, and seasonal workers, plus major employers affected.
206.48(b)(6)Where a state wants something to plan against, the usable proxy is the IHP Cost to Capacity ratio. FEMA publishes it in the IA Declarations Factors Guidance and updates the historical tables annually. It is explicitly not a threshold, but it is the arithmetic FEMA actually runs, and it is why the same dollar loss reads differently in Mississippi than in Connecticut.
Identical $2.4 million in estimated IHP cost, run against each state's fiscal capacity. Virginia's bar sits roughly 4.8 times higher in dollar terms. This is the framework working as designed, not a flaw: capacity is the denominator, so a state that can raise more revenue is expected to absorb more before federal assistance is supplemental.
Switch the table above to the Individual Assistance view to see, for each state, the estimated IHP cost that would put a request in each historical band.
| ICC ratio | Requests received | Approved | Share approved |
|---|---|---|---|
| > 25 | 61 | 57 | 93% |
| 10 to 25 | 90 | 69 | 77% |
| < 10 | 281 | 122 | 43% |
| Total | 432 | 248 | 57% |
The odds moved, and they moved a long way.2019 guidance compared with the 2026 update
| ICC band | 2019 guidance | 2026 update |
|---|---|---|
| Above 25 | 85% | 93% |
| 10 to 25 | 45% | 77% |
| Under 10 | 10% | 43% |
One caution on reading these side by side. The 2026 series is not the 2019 series with nine more years bolted on. The above-25 band holds fewer requests in the longer series than in the shorter one, 61 against 65, which is only possible if FEMA recomputed the whole thing. Treasury revises TTR every year, so a past event's ICC shifts as its denominator shifts and events migrate between bands. Subtracting one table from the other to derive a recent-years figure produces a number that means nothing.
| Estimated IHP cost of assistance | Requests | Declared | Share declared |
|---|---|---|---|
| $7.5M or more | 109 | 103 | 94% |
| $1.5M to $7.5M | 231 | 132 | 57% |
| $1.5M or less | 92 | 13 | 14% |
| Total | 432 | 248 | 57% |
Both tables: FEMA's 2026 update, covering major disaster declaration requests including IHP from January 2008 through 1 December 2025. Expedited requests and those without estimates are excluded. Read these as likelihood in historical context, which is exactly how FEMA frames them.
The order is not administrative trivia. Damage estimates are locked in early by people working fast under bad conditions, and everything downstream is built on them.
Local response begins. Cost capture starts here or it does not happen. Force account labor, equipment hours, contracts, mutual aid. Undocumented response work is unrecoverable work.
Establishes the legal basis for state resources and later counts toward the disaster history FEMA weighs under the cumulative-effects factor.
Where the outcome is decided. The PDA produces the damage figures that every threshold comparison then runs against, and on the IA side it produces the estimated cost of assistance that drives the cost-to-capacity ratio. A weak PDA cannot be repaired later.
Must certify the severity exceeds state and local capability, commit to cost share, and make the case on the factors. This is where a state argues localized impacts, recent disasters, and its own expenditures, the parts no indicator captures.
Regional and headquarters review weigh the factors in totality. The indicators inform this step; by statute they cannot determine it.
Sole authority. The President may decline a FEMA recommendation or declare without one. A denial can be appealed once by the governor, generally within 30 days, and new information is what makes an appeal worth filing.
"We cleared the county threshold, so we get a declaration."
The countywide indicator governs which areas get designated within a declaration under 206.40(b). It does not determine whether a declaration happens at all. That turns on the statewide picture and the factors in totality. A county can clear $4.86 per capita many times over and still sit inside a state that never receives a declaration.
"$1.94 is the cutoff. Under it you get nothing."
It is an indicator, and Section 320 is why. FEMA has recommended PA below it on localized impacts, recent multiple disasters, and mitigation investment, and has declined requests that cleared it. Roughly 80 percent of declarations between 2005 and 2014 met or exceeded the indicator, which also means about a fifth did not.
"IA has a per capita threshold too, we just need to find it."
There is none, and the absence is deliberate. FEMA's own declaration criteria state that no per capita indicator is established as a factor for the consideration of Individual Assistance. Anyone citing an IA per capita figure is either confusing it with PA or repeating something that does not exist.
"The fiscal year figures apply based on when it was declared."
Since FY2025 the notices key to incident start date, not declaration date. An incident beginning 28 September and running into October falls under the prior year's indicators even if the declaration comes months later. Older guidance still circulating says "declared on or after" and is out of date on this point.
"We're at $189 per capita, that's well over the threshold."
Different number, different regulation, different purpose. The $189 figure under 44 CFR 206.47 governs whether FEMA recommends raising the federal cost share from 75 to as much as 90 percent. It has nothing to do with eligibility, and it runs on a calendar year rather than a fiscal year.
"Get a rough damage number in fast, FEMA verifies it later."
The PDA estimate is the number. It feeds the per capita comparison and the estimated cost of assistance, and there is no later step that corrects an undercount. Insurance status matters as much as damage count on the IA side, because FEMA cannot duplicate insurance proceeds. A well-insured county generates less demonstrable IA need from identical damage.
Several efforts to rewrite the declaration framework are in circulation. None has been finalized, none has a Federal Register notice behind it, and none should be planned against. Listed here only so they are recognizable when they come up secondhand.
All statewide figures here use 2020 Census resident population, which is
what state emergency management agencies publish in their own threshold tables. Florida's
Division of Emergency Management publishes a FY25/26 statewide threshold of
$41,784,082.78; the calculation here reproduces that to the penny, which is the
cross-check that the population base and indicator are both right.
FEMA's unfinalized 2020 proposed rule would switch to annual population estimates. If that ever lands, every figure in the table moves.
At least one national association summary cites the FY2026 indicator as $1.97. The figure
is $1.94. The Federal Register notice adjusting the countywide indicator to
$4.86 states the CPI-U increase as 2.9 percent for the 12 months ending August 2025; applying
that to the FY2025 values gives 1.89 × 1.029 = 1.9448 and
4.72 × 1.029 = 4.857, which round to $1.94 and $4.86. State published
tables use $1.94.
The FY2025 and FY2026 notices changed the trigger language. Both indicators now apply to major disasters with an incident start date on or after 1 October, where older notices keyed to the declaration date. For an incident that begins in late September and runs into October, the prior fiscal year's figures govern.
Under 44 CFR 206.47, FEMA uses a separate statewide per capita indicator to recommend
raising the federal cost share from 75 to as much as 90 percent. For disasters declared
1 January through 31 December 2026 that figure is $189 per capita. It governs
cost share, not eligibility, and is adjusted on a calendar year rather than a fiscal year.
Treasury publishes TTR for the 50 states and DC only, so the IA capacity columns are blank for territories; FEMA substitutes GDP by state there, available from BEA for the Virgin Islands, Guam, American Samoa, and the Northern Marianas, and from Census for Puerto Rico.
Tribal governments requesting directly are evaluated under separate criteria in FEMA's Tribal Declarations Interim Guidance, effective for incidents on or after 3 December 2024, not under the state per capita indicator. Those thresholds are not in this table.
The IA view shows 25 × TTR and 10 × TTR expressed in
millions, the estimated IHP cost at which a request would cross into each
band. These are lower edges. An ICC of exactly 25 sits at the bottom of the top band, not
the top of it. Approval percentages come from FEMA's 2026 update, retrieved from the agency
directly; both its tables independently total 432 requests and 248 approvals, and all six
percentages recompute from their own row. TTR is the 2023 estimate from Treasury's September 2025 release, the most
recent published. Treasury data runs about two years behind, which FEMA acknowledges; a state
whose capacity dropped recently should say so in the request and point to GDP instead.
These are not thresholds. No ICC value entitles a state to a declaration, and requests have been approved well below the bands and denied well above them.