The End of Pandemic Aid: What Two States’ Spending Records Reveal

The signal

Grant exhaustion is a funding-source event. Its effect on services, contractors and credit depends on project timing and enacted replacement decisions.

The deadline is precise; the economic shock is not

Treasury distinguishes the December 31, 2024 obligation deadline from the general December 31, 2026 expenditure deadline. Surface Transportation and Title I eligible-use projects had a September 30, 2026 expenditure deadline. The September newsletter also points to rules for closeout costs, invoice processing and checks extending into 2027. Those administrative provisions do not amount to a blanket extension for new spending. [1]

An obligation commits funds under the program’s rules. An expenditure records a qualifying use or payment under the relevant reporting framework. Neither is the same as an annual recurring budget line. A completed infrastructure project, a temporary household payment and an ongoing service can all draw from the same grant program while producing very different fiscal needs after it ends.

The right question for 2027 is therefore narrower than the original award: which services are intended to continue, what is their annual cost, and what recurring funding has been enacted to pay for them? A deadline can accelerate project completion and compliance work without causing an equal reduction in the next year’s payroll or service spending.

The national snapshot is substantial but dated

GAO’s February 2026 update reported that states had spent US$156.3 billion, or 80% of their awards, as of March 31, 2025. Localities had spent US$107.2 billion, or 84%. Revenue replacement accounted for US$82.6 billion of state spending and US$71.9 billion of local spending. The underlying snapshot predates the current deadline debate by more than a year and a half. [2]

These figures establish the historical scale and broad uses of the program. They do not establish what remains in November 2026. Subtracting that old expenditure figure from original awards and calling the remainder a current fiscal shock would combine a stale balance with an unsupported interpretation.

Even a fully current balance would still need classification. Revenue replacement is a reporting category with broad uses; it is not proof that every dollar is financing a permanent staffing commitment. The state and local aggregates also have different institutional coverage. Our document comparison keeps them as context and does not merge them into the two-state observations.

California’s cumulative total is already near the finish line

California’s September table reports US$27.016678 billion in cumulative expenditures against a US$27.017017 billion final allocation, with US$0.338 million shown as returned funds. Its revenue-replacement category accounts for US$22.482578 billion. The table is in thousands of dollars; converting the units is essential before comparing it with budget documents. [3]

The June table reported US$27.016707 billion of cumulative expenditure and the same US$22.482578 billion revenue-replacement amount. Between the two tables, total reported expenditure fell by US$29,000 while returned funds increased by the same amount. That small adjustment is evidence of reporting and closeout mechanics, not a meaningful quarterly contraction in California’s economy. [4]

This example sharply limits what the federal deadline can tell us. A program already recorded as largely expended cannot be treated as though its full multiyear award is supporting one final annual spending flow. Any continuing fiscal pressure must be traced through the activities previously financed and their replacement budgets. The grant total alone supplies neither the annual cost nor the amount still unfunded.

Cash advances complicate the link to activity on the ground

California explicitly says its expenditure totals include cash advances to subrecipients for goods or services that have not yet been invoiced. Thus, a state-level expenditure record can precede the ultimate delivery of a service or completion of work. The tables also identify abatements and reclassification in their returned-fund notes. [3] [4]

This does not make the reported amounts unreliable; it defines what they measure. An analyst studying state treasury cash movements can use them directly on that basis. An analyst forecasting construction activity or household services needs the next stage of the chain: subrecipient spending, contracts, delivery schedules and applicable program rules.

The same distinction helps contractors. A state may have transferred funds while a project remains active, or a project may be complete while final accounting adjustments remain. Neither the appearance of a zero remaining balance nor the arrival of a federal reporting deadline proves that new invoices stop on the same day. Contract terms and project status remain the operational evidence.

Pennsylvania shows why the project mix matters

Pennsylvania’s general revenue-replacement project, PA1, recorded cumulative expenditure of US$4.620290 billion at September 30, compared with US$4.619314 billion in June. The third-quarter flow was US$975,894. Its September obligations exceeded cumulative expenditure by about US$4.723 million. These are records for one large project, not the whole state award. [5] [6]

The contrast between a US$4.6 billion historical total and a quarterly flow below US$1 million is the central point. Treating the total as an annual service subsidy about to disappear would radically misstate the timing. The program description says it counteracts revenue loss and supports government services, but it does not identify a matching recurring 2027 funding gap.

The same report identifies a US$140 million housing-subsidy project as a one-time bonus and completed. A temporary benefit can end as designed without leaving a standing obligation to repeat it. That is a substantive counterexample to the assumption that all expiring grant expenditure must be replaced indefinitely. Its ending can still matter to recipients, but the budget classification should preserve its temporary design. [5]

Water projects still have a measurable cash pipeline

Two Pennsylvania infrastructure projects have a different pattern. Water & Sewer Projects recorded US$62.182286 million of cumulative expenditure at September, up from US$45.444178 million in June. H2O PA rose to US$119.390502 million from US$75.842627 million. Together, their third-quarter expenditure was US$60.285983 million. [5] [6]

Subtracting cumulative expenditure from the two projects’ September obligations leaves US$136.309371 million. This is a remaining project funding amount on the report’s basis. It is not an annual recurring-service deficit, a forecast that all of it will become fourth-quarter construction output, or evidence that the state lacks resources to complete the work.

The economic channel is nevertheless concrete. Project execution, reimbursement and completion can support a temporary payment flow for contractors. Once a project finishes, that flow may decline even if ordinary government services remain intact. A business exposed to this pipeline should inspect contract completion and replacement orders; a municipal-credit analyst should separately examine operating-budget obligations. The two exposures need different tests.

The replacement calculation must start with an annual flow

Our calculator uses documented annual recurring grant-funded expenditure that is ending, minus enacted recurring replacement funding. A positive result is an uncovered modeled annual requirement. A zero or negative result means no uncovered requirement under those inputs. Unspent grant balances, cumulative awards and proposed measures do not belong in the recurring-funding fields.

For a hypothetical state program costing US$300 million annually, US$180 million of enacted recurring replacement leaves US$120 million to fund each year. If US$240 million of reserves is genuinely available and uncommitted, it provides two years of arithmetic coverage at unchanged spending and revenues. That is a temporary buffer, not a recurring solution or legal permission to spend the reserves.

If enacted replacement rises to US$300 million, the modeled gap closes. The tool reports no uncovered gap and suppresses reserve coverage rather than displaying infinity. If only US$60 million of reserves is legally available, the original gap has six months of coverage. These scenarios describe the sensitivity of a budget bridge; they are not calibrated estimates for California or Pennsylvania.

What would strengthen or overturn the cliff thesis?

The strongest evidence for a recurring cliff would identify a continuing service, show its expiring annual grant support, and demonstrate that an enacted budget leaves it unfunded or relies on reserve depletion. Evidence of service cuts, delayed payments or replacement taxes would then connect the funding change to households and businesses.

The countercase is equally testable. A state may have used grants for completed projects, temporary transfers, or services already moved onto recurring revenues. In that case, the headline grant expiration overstates the direct recurring-budget shock. The remaining risk may instead concern project deadlines, closeout compliance or the decline in temporary contractor work.

Our evidence supports a differentiated watchlist rather than a state credit ranking. California’s records show how little a near-complete cumulative total says about the next annual budget. Pennsylvania demonstrates that a largely spent revenue-replacement project can coexist with active infrastructure payments. The practical decision is to follow the relevant program into its fiscal-year budget and delivery contracts, keeping the remaining grant stock, current payment flow and recurring replacement requirement in separate columns.

The evidence, side by side

Recurring Grant Replacement Bridge

Hypothetical annual budget bridge; no state-specific forecast, legal authorization or default probability.

Finite nonnegative inputs on consistent fiscal-year and currency bases. If no positive requirement, report no uncovered modeled gap and reserve coverage unavailable. Grant stocks must not be entered as annual service flows.

Assumes unchanged annual requirement and reserve availability. Reserves do not become recurring revenue. Planned funding must not be labelled enacted.

Scenarios to watch

Partial recurring replacement

US$300m annual cost less US$180m recurring replacement leaves US$120m; US$240m available reserves cover 24 months arithmetically.

Full recurring replacement

US$300m replacement leaves no uncovered modeled gap; coverage is not reported as infinity.

Restricted reserves

Only US$60m available against the original US$120m requirement provides six months of arithmetic coverage.

Method and limitations

Two-state document study using June and September 2026 expenditure reports, Treasury’s September guidance and GAO’s dated national review. Six primary documents were read by 13 November 2026. California totals retain their cash-advance definition. Pennsylvania comparisons match project IDs across reports. Calculations distinguish cumulative expenditure, quarterly flow and obligations less expenditure. No 50-state panel, enacted-budget replacement ranking, tax elasticity or 2027 default model has been collected.

The reports do not by themselves identify an annual recurring replacement gap or legally available reserve balances for either state. Calculator inputs are hypothetical.

Publication dates of the state PDFs are not displayed; their observation periods are explicit, and the documents were publicly accessible by the research cutoff.

Reported expenditure can include advances, revisions and reclassification; it is not identical to contemporaneous service output.

Primary sources and evidence register

Source dates and reporting periods differ. Several documents from one issuer are one source family, not independent confirmations. The evidence download contains the observations used in this article.

Frequently asked questions

Does every state lose its whole award in 2027?

No. Awards are multiyear stocks and much has already been spent. The recurring issue is the annual cost of services intended to continue after their grant support ends.

Is an obligation the same as expenditure?

No. The program distinguishes commitments from qualifying expenditure. Completion, invoices and closeout can have different reporting dates.

Can reserves solve a recurring funding gap?

They can provide temporary coverage if legally available and uncommitted. They do not create recurring revenue, and the coverage ratio is not a default probability.

Does fully reported expenditure mean all work is complete?

Not necessarily. California explicitly includes cash advances to subrecipients. Project delivery and final expenditure compliance require separate evidence.