Disaster recovery block grants received a statutory authorization on 2026-07-11 that terminates on 2029-07-11 — the same Act bars authorizing any additional appropriations

What is happening?

The United States has put more than USD 111 billion into Community Development Block Grant Disaster Recovery since fiscal year 1993, roughly USD 65 billion of it since fiscal year 2016, and until this summer every dollar ran through a program that no statute had created. A Congressional Research Service brief dated 2026-05-08 states the arrangement flatly — CDBG-DR is not a program with its own standing authorization or regulations. HUD administered it under the ordinary block grant statute plus whatever waiver power each individual supplemental appropriations act happened to grant.

That changed on 2026-07-11, twenty-eight days before this document was checked. The 21st Century ROAD to Housing Act became Public Law 119-101 without a presidential signature, and its Section 504 adds a new Section 124 to the Housing and Community Development Act of 1974, codified at 42 U.S.C. 5324. A statute establishing the program now exists.

Four things it did not do are readable in its own text. The program terminates on 2029-07-11, three years after enactment, where the standing recommendation from the Government Accountability Office had been permanent authorization. Section 504 creates a Treasury account called the Long-Term Disaster Recovery Fund, and Section 1202 of the same Act then bars authorizing any additional appropriations to carry out the Act or any amendment made by it. Section 504(i) confines the new section to money appropriated after enactment, so the existing portfolio keeps running under the old notice regime. And grants are limited to catastrophic major disasters, a term HUD must define by regulation, with final rules not due until 2027-07-11.

The delay the reform was aimed at is measured. For disasters from January 2017 through January 2023, the interval from declaration to congressional appropriation ran from over 2 months to nearly 2 years, and the interval from appropriation to HUD allocation ran from over 4 months to over 3 years, across USD 49.4 billion in 104 grants to 55 grantees, as of 2025-09-29.

Whose problem is this?

The body that can end the waiting is the same body that has to be asked, disaster by disaster, to start it. Nothing in the new statute changes which body that is.

Where does this problem end?

The boundary matters here because the money was never the missing thing. Congress appropriated it more than thirty times over three decades. What was missing was a program to receive it, and the reform that arrived created one whose account the same law forbids authorizing money for.

What is the state now, and what should it be?

Now

Needs a new measurementthe target state: no source opened here names how fast recovery money is supposed to move, how much the Long-Term Disaster Recovery Fund is supposed to hold, or what share of declared disasters the standing program is meant to reach. GAO publishes intervals and calls the system vulnerable, but publishes no benchmark against which any interval is late. Note also that two statutory dates in the table above lie in the future and are therefore not missed — the formula notice is due 2026-08-10 and the final regulations are due 2027-07-11.

How big is it?

The January 2025 allocation notice states the method in its own text — unmet housing need is the number of housing units with unmet needs, multiplied by the estimated cost to repair those units, less the repair funds estimated to come from FEMA, the Small Business Administration and insurance. HUD therefore holds a count of housing units. The notice reports the result to the dollar, USD 10,338,640,000 of unmet need plus USD 1,550,797,000 of mitigation, and reports no count of housing units, households or people anywhere in the document. The product cannot be inverted, because it covers housing, economic revitalization and infrastructure together and neither the housing share nor an average per-unit repair cost is published.

Three further inputs are absent. There is no published conversion from housing units to persons. There is no deduplication across grants, and a household in a repeatedly declared county can fall inside several allocation rounds. And there is no forward-looking denominator at all, because the new program reaches only catastrophic major disasters and the criteria for that word do not exist until the final regulations. Scope indicators do exist and must not be read as population counts — about 50 grantee jurisdictions received allocations in the January 2025 round, 55 grantees held 104 grants for disasters declared between January 2017 and January 2023, and 27 disasters in 2024 caused at least USD 1 billion in damages each. Summing the populations of grantee jurisdictions would count every resident of a state as affected by a single county allocation, which is exactly the inference the targeting standard exists to reject.

Needs a new measurementthe affected population is not derivable, and the reason is itself the finding. HUD publishes the product of its own formula and withholds the multiplicand.

Under what conditions does it arise?

1. No standing appropriation means every disaster begins by asking Congress. Because the program had no money of its own, recovery funding travelled as a supplemental appropriation, which is a bargaining vehicle. Its timing therefore tracked unrelated legislative business rather than the damage, and the measured interval from declaration to appropriation ran from over 2 months to nearly 2 years.

2. The appropriation was the authorizing instrument. With no program statute, the appropriations act itself had to hand HUD the power to waive or specify alternative requirements for any provision of any statute or regulation the Secretary administers. The rulebook was created by the money, so it logically could not precede it. This is the structural core of the whole thing.

3. The rules were then written in public, once per grant. GAO counted more than 100 Federal Register notices tied to disaster recovery funds since 2001 and tied the resulting variation directly to grantee difficulty in managing several grants at once. For the 2017 hurricanes, 5 months elapsed between the first appropriation and the first HUD notice establishing requirements.

4. Execution runs years behind allocation. Grantees submit an action plan, HUD approves it, and only then does procurement and construction begin. Of the roughly USD 3.7 billion that grantees budgeted to satisfy FEMA cost-share obligations, about 8 percent had been expended as of December 2024.

5. The reform has no vehicle of its own. Its constituency is geographically concentrated and episodic, so it moves only when attached to something larger. In 2025 it was attached to the National Defense Authorization Act and dropped from the House-passed conference bill on 2025-12-10.

What has been tried?

Two directions were tried at once — legislate a standing program, and compress the administrative step without one. The second worked and is visible in the 26-day interval. The first took from the 116th Congress to 2026 and arrived with a termination date.

What was found?

Why is it still unsolved?

Institutional gap — the institution now exists in statute and lacks the two properties that would end the per-disaster reconstruction, which are permanence and an endowed account.

Start with the pairing that decides everything else. Section 504(c) establishes the Long-Term Disaster Recovery Fund in the Treasury and says it shall consist of amounts appropriated, transferred and credited to it. The only transfer sources it names are recaptured or unobligated balances from prior disaster appropriations under the Community Development Fund heading, which is money already inside the system rather than money entering it. Then Section 1202, the final general provision of the Act, forbids authorizing any additional appropriations to carry out the Act or any amendment made by it. Within a single law, disaster recovery received a Treasury account with no money line and three years, while the Innovation Fund in Section 208 received USD 200 million a year for five years and a seven-year horizon. No opened secondary source mentions this pairing at all.

The second movement is that the new clocks sit downstream of the old one. Section 124(a)(6) gives HUD 90 days after a declaration to determine that a disaster is catastrophic and qualifies, extendable to 120 days where data is insufficient, and Section 124(a)(4) requires a formula methodology notice within 30 days of enactment. These are real improvements to the two steps HUD controls, and HUD had already shown those steps were compressible without any statute — the Universal Notice took one appropriation to allocation in 26 days against a historical range of 4 months to over 3 years. But the sentence that carries the machinery is conditional. HUD shall immediately announce an allocation if amounts are available in the Fund, and otherwise within 15 days of the appropriation. The sequence therefore still begins where it always began, with Congress deciding disaster by disaster to put money in.

The third movement is durability, and it is where the arithmetic stops closing. The authorization runs three years and carries a sense of Congress that the sunset shall not preclude a successor, which is the statute stating in its own text that it expects to be replaced. The final regulations are due one year in, so the program operates under finished rules for at most two of its three years, and the eligibility threshold does not exist during the first year. A grantee spend-down window of six years sits inside a three-year program, so a grant obligated near the end runs years past the termination date. No source opened here addresses that overhang, or explains why a recommendation for permanent authorization was answered with three years. The reform converted a program that did not exist until Congress acted into a program that cannot pay until Congress acts.

What observation would mean it is solved?

Candidates — (a) an appropriation or transfer actually lands in the Long-Term Disaster Recovery Fund and HUD announces an allocation on the statutory determination clock rather than after a supplemental (b) the interval from major disaster declaration to HUD allocation announcement falls and stays low across several consecutive disasters (c) Congress removes the 2029-07-11 termination date or replaces the section with a permanent authorization.

(a) alone proves less than it looks. A single deposit could be small, one-time, or drawn entirely from recaptured balances of older grants, which moves money around inside the system without adding any. The statute permits exactly that, and no source states how much the Fund would need in a year to matter.

(b) alone can move for reasons that have nothing to do with the reform. The 26-day interval in January 2025 was produced administratively, before any statute existed, and it was one appropriation. A quiet disaster year lowers the interval, a contested supplemental raises it, and the catastrophic threshold has not been written yet, so the set of disasters that will even enter the count is undefined.

(c) alone counts paper. Permanence without an endowed account reproduces the present arrangement with a longer horizon. The three candidates have to be read together, and (b) has to be read against the year it came from and against how narrow the catastrophic definition turns out to be.

What is it connected to?

Fills with researchFEMA Individual Assistance and Public Assistance, the Small Business Administration disaster loan program, the National Flood Insurance Program, coordination across the roughly 30 federal agencies GAO counts in disaster recovery, and the competing proposal to house a dedicated recovery fund at FEMA rather than HUD. Relation type and evidence grade were not confirmed in this round.

What these sources do not say

• HUD publishes the product and deletes the multiplicand. The allocation formula is a housing-unit count multiplied by a repair cost, less other federal and insurance funds. The January 2025 notice reports the dollar result to the unit and never reports the housing-unit count anywhere in the document. No source states why the count is withheld or where it is published.
• The most-cited evidence for the problem is a number nobody defines. The count of Federal Register notices tied to disaster recovery appears as more than 100 since 2001 in one source, more than 80 since 2001 in another, and 145 term matches in a 2026 database query. Whether a count includes allocation notices only, or waivers, or every document mentioning the term, is stated by none of them.
• Deadlines ran one direction until 2026. The Universal Notice binds grantees to submit an action plan within 90 calendar days. No corresponding deadline binding HUD to publish an allocation notice after an appropriation was found in any source opened here, and the new statutory clocks apply only to money appropriated after 2026-07-11.
• No source states how fast recovery money should move. GAO publishes intervals of 2 months to 2 years and 4 months to over 3 years and offers no benchmark. Without a target, a three-year appropriation-to-allocation interval cannot be called late by any published standard.
• Nothing explains why the fix is three years. The recommendation asked Congress to consider permanent authorization. The statute gives three years plus a sense-of-Congress clause anticipating a successor, and no opened source explains the departure or addresses the six-year grantee spend-down window sitting inside it.
• The provision that decides whether the program can operate is discussed nowhere. Not one opened secondary source mentions Section 1202 or the availability condition in Section 124(a)(6)(B). None states how much the Fund would need annually, or relates it to the 27 billion-dollar disasters counted in 2024 alone.
• Whether anything is in the Fund is unpublished. A Federal Register query on 2026-08-08 returned zero documents mentioning the Long-Term Disaster Recovery Fund. That is absence of evidence, not evidence of an empty account, and the first statutory notice is not due until 2026-08-10.
• How narrow the catastrophic threshold is has not been estimated by anyone, and the fate of the existing portfolio is unstated. No source estimates how many past disasters would have qualified as catastrophic, which is the central question about whether the standing program is narrower than the appropriations practice it replaces. Nor does any source state how many grants remain open under the old regime or when they close.

See the evidence

The statute was read directly and everything structural rests on it. The public law was retrieved as a government publishing office PDF and extracted locally, and every section reference above comes from that text rather than from any summary. Two Federal Register notices were read in full through the full-text endpoint, which is why the absence of a housing-unit count in the allocation notice can be asserted rather than assumed. Everything else is secondary. Where sources disagree the disagreement is left visible — one issue brief says the Act was signed into law while the note printed in the public law itself states it became law without a presidential signature, and the primary text controls; one third-party explainer gives the bill number as H.R. 6344 against H.R. 6644 in the enrolled text and is treated as unreliable for this problem; and the count of Federal Register notices stands at three different values from three different bases with no counting rule published by any of them. Five rows could not be opened and are kept above with their reasons rather than replaced by a parent agency homepage. This is a Path A output, so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range Not derivable

The reason and what is missing are listed under “What is missing” below