Study: Extra ESSER aid near a Title I cutoff got offset by local tax cuts

A new NBER working paper is sharpening one of the hardest lessons from the ESSER era: for school districts sitting near a key Title I cutoff, extra federal pandemic aid appears to have been neutralized before it showed up as additional classroom spending. In research featured by NBER on July 30, 2026, economists Jeffrey Clemens, Philip G. Hoxie, and Stan Veuger found that districts just above a 5 percent poverty threshold received about $390 more per student on average, yet did not increase per-pupil expenditures or post measurable math and reading gains through 2023 because local revenue collections, including property taxes, fell alongside the added federal money.

That finding matters in August 2026 because districts are operating after the main ESSER deadlines: ARP ESSER funds had to be obligated by September 30, 2024, and the Education Department later tightened liquidation rules, while allowing only narrower project-specific extensions. In other words, school systems are now living with the post-ESSER budget reality, and this paper helps explain why a large federal appropriation does not automatically translate into more staff, tutoring, or instructional time on the ground. (ed.gov)

A study at the formula margin, not a verdict on all ESSER

Congress sent roughly $189.5 billion to K-12 schools through three ESSER rounds, and the Education Department distributed those funds to states using Title I formulas. The new paper focuses on a specific feature of those formulas: districts became eligible for Title I Targeted Grants and Education Finance Incentive Grants once their formula-child rate reached 5 percent. Because ESSER allocations were built on Title I distributions, districts clustered just above and just below that line received different federal windfalls for reasons that were largely mechanical, giving the authors a plausible way to isolate the effect of the extra money. (ed.gov)

The authors use what economists call a difference-in-discontinuities design. In plain English, they compare districts near the same threshold before and after the pandemic aid arrived, rather than comparing obviously different districts such as affluent suburbs and high-poverty urban systems. Their preferred estimate is that crossing the cutoff generated about $388 in additional ESSER aid per pupil, essentially the same as the “about $390 per student” figure highlighted in NBER’s July 30 summary. That is a meaningful funding bump, but it is still a marginal estimate for districts near one eligibility line, not a measure of the effect of every ESSER dollar nationwide. The paper is also a working paper, not yet peer reviewed. (nber.org)

Why the extra dollars did not show up as extra spending

The paper’s central result is not that districts got no money. It is that the incremental federal money did not stay inside district budgets as additional education spending. On the authors’ account, the aid was “passed on” through lower local revenue collections, including property taxes, leaving no increase in per-pupil expenditures and no detectable improvement in math or reading scores. For school finance readers, that is a classic fungibility story: when one layer of government sends more aid, another layer can pull back. (nber.org)

That conclusion sits awkwardly, but not necessarily inconsistently, beside the broader federal record on ESSER spending. A 2024 Government Accountability Office review found that districts nationally reported spending nearly $60 billion in ESSER funds through school year 2021-22, with about 80 percent going toward students’ academic, social, and emotional needs and the continuation of school operations. Those national figures show that districts did spend ESSER dollars. The new paper asks a narrower question: whether the extra dollars generated by one formula notch produced extra education spending after local fiscal responses were taken into account. (gao.gov)

Federal policymakers were not blind to that risk. The Education Department’s maintenance-of-equity guidance explicitly warned that ARP ESSER’s targeting could be undermined if state or local funds were cut, and ARP required districts to reserve at least 20 percent of their allocation for learning-loss interventions. But those safeguards were aimed largely at protecting the highest-poverty systems and schools from disproportionate cuts, not necessarily at preventing every form of fiscal substitution around the margins of local taxation. Our reading of the new paper is that future rescue packages may need tighter supplement-not-supplant rules, more school-level reporting, or narrower allowable uses if lawmakers want emergency dollars to change classroom conditions rather than relieve pressure elsewhere in local budgets.

What readers should — and should not — take away

The easiest misuse of this paper would be to treat it as proof that ESSER “failed.” That goes further than the evidence supports. The authors study districts near a specific 5 percent threshold and outcomes through 2023, emphasizing spending, local revenue, and test scores. The same paper also reports suggestive evidence of greater political engagement, stronger enrollment retention, and faster reopening in districts that received the extra funds. Those are not trivial outcomes, even if they do not settle the learning-recovery question. (nber.org)

It also matters that other researchers, using a different design and a much broader sample, have reached a more positive conclusion about ESSER’s academic effects. A 2025 CALDER working paper examining nearly 5,000 districts in 28 states estimated that each $1,000 increase in ESSER spending per pupil modestly improved math achievement in 2023, while producing no statistically significant gain in English language arts. Read together, the two studies suggest a more precise lesson for school leaders and lawmakers: federal aid can help when it actually adds to school spending, but the effect is smaller or disappears when local finance systems absorb the windfall.

That is the practical takeaway for superintendents, chiefs of finance, and state officials now planning for the next downturn or disaster. If Congress wants emergency school aid to buy tutoring slots, preserve staffing, extend learning time, or stabilize student supports, lawmakers may need stronger guardrails than a poverty-weighted formula alone. If they do not, the next rescue package could again be large on paper but surprisingly faint at the classroom level. As states and districts move farther from the pandemic and deeper into the ESSER afterlife, that design question may matter more than the next topline appropriation number. (nber.org)