OCG White Paper No. 015 — Has Ohio Been Defunding Public Education?

Source. Ohio Common Ground Research Center, “Has Ohio Been Defunding Public Education? A Two-Decade Analysis of School Funding, Enrollment, Spending, and the Claims Made About Them — With Selected Updates Through FY2025.” White Paper No. 015, Tracking ID RL-2026-023, Version 1.0, August 11, 2026. Type. Secondary source — analysis over primary data. Location. ohiocommonground.com.

What it contains. Five commonly repeated claims about Ohio school funding, each tested measure by measure against four deliberately separated funding “universes” (all-funds NCES, state appropriations, foundation-delivered, scholarship payments). Its central device is a table of seven standards for the word “defunding” — nominal reduction, real reduction, per-pupil reduction, declining state share, purchasing-power erosion, unfunded statutory formula, and funding relative to obligations — with Ohio’s answer under each.

This is the strongest of the three OCG publications this corpus holds. Keeping the universes apart and never combining them, assigning a verdict per measure rather than collapsing to a grade, and returning Insufficient Evidence on the “bankrupting” claim rather than stretching to an answer are all correct choices. Section 8 also avoids the error this corpus recorded against White Paper 013: it reports statewide NAEP alongside funding with an explicit no-causation caveat rather than running a district cross-section.

Full corroboration on the Auditor series. Every figure it draws from the Ohio Auditor’s Longitudinal School Finance Study matches this corpus’s independently held values, which come from the same source through the deflator calculator: $7,065 → $15,314, +116.8% nominal and +26.1% real; $14,493 excluding relief, +19.4%; total operating expenditure $12.97B → $25.78B; enrollment −8.3% from 1.84M to 1.68M; CPI +71.9%; instruction +25.7% and support services +28.2% real. verified see crates/deflator/tests/ohio_epp_real_series.rs

Three findings the corpus adds # ↑ Contents

1. The real series is not monotone, and the endpoint framing conceals a real decline. Figure 1 plots the nominal series, which rises smoothly. Deflated year by year, real per-pupil spending fell about 7% from FY2010 to FY2014 — $15,226 to $14,173 in constant FY2022 dollars — leaving FY2014 indistinguishable from FY2006 and not recovering to the FY2010 level until FY2018. verified

The seven-standards table answers “real reduction” and “per-pupil reduction” with No. That is correct for FY2000→FY2022 and false for FY2010→FY2014, which is plausibly the period a “they defunded us” speaker has in mind. The paper’s structural thesis — that the disagreement is definitional — is right and incomplete: it is also periodic. Two speakers can choose different windows on one series and both be correct, and no column in the table surfaces that.

The interior rows rest on chart labels read to about $100; the decline is $1,053, and a test perturbs both endpoints in the direction that would erase it and finds it intact.

2. Real per-pupil spending peaked in FY2020, not FY2022. $15,747 against $15,314 — the FY2022 endpoint sits $433, or 2.7%, below the real peak, because relief arrived while prices were low and was then eroded by FY2021–22 inflation. verified The “record funding” verdict is marked Supported on the real per-pupil row; that is true against FY2000 and not true as a record.

3. At full phase-in the “unfunded statutory formula” standard does not resolve — it inverts. Section 5 reports FY2024: $741.8M below the formula, 180 districts drawing $181.2M in guarantee. The department’s own FY2027 model — the terminal year, formula at 100% — shows 294 of 609 districts (48.3%), holding 54.1% of Ohio’s students, funded by the guarantee rather than the formula, totaling $878,974,300. verified Roughly fivefold growth in guarantee dollars arriving at completion of the phase-in.

The guarantee is anchored to FY2020 and has never been re-based; for guaranteed districts the FY2027 formula produces a median 67.8% of that baseline. The paper’s Future Research asks to “track FSFP phase-in through FY2027 and whether the formula is funded to its calculated amount.” Full phase-in does not put districts on the formula; it leaves half the state on a hold-harmless anchored to a year the formula was not run.

The trough, now located district by district # ↑ Contents

The paper’s per-pupil-reduction row reads No for FY2000-FY2022. Over FY2010-FY2013 the district panel shows 81% of Ohio’s comparable districts spending less per pupil in real terms, median -5.6%. verified It was not a few districts moving an average, and nothing about a district — its state, federal or local revenue share, its later poverty or property wealth — predicts how far it fell. A general contraction, not a targeted one.

Extending the same panel back to FY2009 also enlarges the paper’s Section 4 finding on its own terms. The state share of district revenue falls 11.7 points from FY2009 to FY2022 on the F-33 comparable-district basis, against the 6.2 points the paper reports for FY2000-FY2022 on the wider all-funds basis. Different universes, not directly comparable — but the direction is the paper’s and the recent decline is steeper than a two-decade endpoint pair conveys. verified

The window it could not reach, and it points the other way # ↑ Contents

The all-funds series ends at FY2022 and Future Research asks for an extension. This corpus holds actual district receipts for 660 districts, FY2020–FY2025, over which CPI rose 25.1%: verified see crates/project/tests/finances_and_the_guarantee.rs

	nominal up	real up	median real change
unrestricted state aid	491/658	196/658	−11.4%
total revenue	637/659	259/659	−2.8%
total expenditure	632/658	380/658	+3.4%
property tax	637/659	345/659	+0.9%

Aggregate unrestricted state aid: $7.82B → $7.89B nominal (+1.0%), −19.3% real, with 70.2% of districts losing real ground. On the paper’s own “real reduction” standard applied to state aid over the six most recent observed years, the answer is Yes. What districts spend held its real value; what the state sent them did not, and the difference was made locally.

The mechanism is recorded on the guarantee node: it is a nominal floor, so a district it protects loses purchasing power annually by construction, with nothing in the formula recording that it happened.

Corroboration on the voucher mechanism # ↑ Contents

Section 6’s load-bearing claim — that post-FSFP scholarships are direct state payments rather than deductions from resident-district aid — is independently confirmed. This corpus searched district payment reports for a deduction channel and found none; the transfer lines are too small and run in both directions. verified see crates/project/tests/the_voucher_channel_is_absent.rs

One standing caution from skills/deduction: a series spanning the FY2022 transition without marking it shows foundation payments rising for reasons unrelated to the formula. The paper’s own comparisons are post-FSFP on both sides and clean; a reader spanning FY2000–FY2025 will not necessarily be.

Section 9, now half-checked # ↑ Contents

The FY2024 F-33 survey is held and the cross-state table splits in two. verified see crates/dispersion/tests/census_f33_fy2024.rs

The revenue mix reproduces, and it is the part the argument rests on. Ohio’s shares come out state 35.1% / local 53.3% / federal 11.6% against the published 34.6 / 53.6 / 11.8; Indiana and Kentucky land within half a point on every line. Ohio’s local reliance sits more than nine points above the national figure and above all three of its western and southern neighbors. The paper’s Section 9 claim is confirmed.

The per-pupil dollars do not reproduce, and the corpus cannot say why. Aggregating the individual-unit file gives Ohio $16,610 against a published $17,257, and the errors scatter rather than shift: Indiana +0.5%, Kentucky +1.1%, Pennsylvania +3.2%, Ohio −3.7%, Michigan −9.8%. A uniform bias would point at a spending definition or a unit filter and could be corrected; a state-specific one cannot be from these rows. The Bureau publishes its state per-pupil figures from summary tables built on something the unit file does not carry.

That is recorded as a limit on this corpus, not as an error in the paper, which quotes a published Census figure correctly. It does mean nobody should rebuild a state per-pupil comparison from the unit file and expect the published numbers. open

What this corpus still cannot check # ↑ Contents

NAEP (Section 8), the FY2014–FY2025 scholarship payment series (Section 6), and the LSC appropriation history remain outside the corpus’s coverage and are recorded as unverified rather than accepted.