Ohio School Report Card

One number, three consequences. That is the fact this class exists to make sayable. R.C. 3302.03 produces an overall rating; R.C. 3302.10 turns three bad ones into the loss of a district’s budget authority; the Fair School Funding Plan turns a good one into money. A corpus that models only the third describes a payment and misses that the same measure carries a sanction an order of magnitude larger in consequence. inference

The scale changed in 2022-23, from letter grades to stars, and both survive in current statute: R.C. 3302.10 states its trigger as an overall grade of “F” under division (C)(3) or an overall performance rating of less than two stars under division (D)(3). verified A series that crosses that boundary is comparing two scales, and the statute itself carries both because the transition is not finished in law.

The department must publish preliminary data before the report card itself. R.C. 3302.03 requires preliminary report card data to districts not later than the thirty-first of July each year. verified

The rating is built from six components and none of them is scored on a fixed standard. The section names the components and weights them, delegates every threshold to administrative rule, and scores achievement against a maximum refitted annually — so a rating is a position in a distribution rather than a level. What that costs is measured below. verified computed; see crates/dispersion/tests/the_weights_behind_the_index.rs

Whether the 2022-23 change of scale moved any district across the distress trigger is not established and is exactly the kind of question the two-regime model exists to ask. open

What this repository computed # ↑ Contents

Not what Ohio publishes. Figures derived here from committed fixtures, each one citing the test that reproduces it.

Six components, weighted as proportions rather than percentages # ↑ Contents

R.C. 3302.03(D)(3) groups the report card’s measures into gap closing, achievement, progress, graduation, early literacy, and college/career/workforce/military readiness. Division (D)(3)(g) then builds the overall rating out of them without ever printing a percentage: achievement and progress carry equal weight, and each remaining component carries half of achievement’s. That is 2:2:1:1:1 under (g)(i) — 28.6% each to achievement and progress and 14.3% to each of the other three — or 2:2:1:1:1:1 under (g)(ii), which is 25% and 12.5% and applies only where the Joint Committee on Agency Rule Review approved the department’s readiness rules. verified

Which of the two governs the 2024-25 report card is not established here. Division (D)(3)(f) makes it turn on a single administrative event — the committee approving or disapproving the rules the department was directed to propose — and the difference is whether the overall rating has five components or six, so it moves every district’s rating. The department’s own achievement download carries the achievement component alone and cannot answer it; the committee’s docket can. open

The star thresholds are not in the section, and their absence is the design # ↑ Contents

Division (D)(4)(a) delegates the “performance criteria, benchmarks, and rating system” to rules adopted under Chapter 119. What the statute sets instead is a shape: (D)(4)(b) requires the department to use prior years and simulations so that there is “meaningful differentiation” and that, outside one named exception, “more than half of all districts or buildings do not earn the same performance rating in any component or overall”; division (M) requires the benchmarks be reviewed at least every three years. The cutpoints are meant to move. verified A star rating is a position in a distribution the department is under a duty to spread, and not a level. Anyone reading one as a fixed standard is reading the wrong kind of number.

The achievement component is scored against the top of the state, refitted every year # ↑ Contents

R.C. 3302.03(D)(1)(c) does not award the achievement rating on the Performance Index. It awards it on the Index divided by a maximum, and defines that maximum as “the average of the highest two per cent of performance index scores achieved by a district for the school year for which a report card is issued.” One number for the whole state, recomputed annually. verified

The department publishes it. For 2024-25 it is 109.8, identical on all 607 district rows. The statute says “two per cent” and does not say how to round it; two per cent of 607 is 12.14, and the file decides: the mean of the top 13 scores is 109.81 and rounds to what was published, while the top 12 give 109.9 and the top 14 give 109.7. The department takes the ceiling. verified computed; see crates/dispersion/tests/the_weights_behind_the_index.rs

So the rating is a share and the Index is a level, and they can move in opposite directions. The denominator was 108.8 in 2022-23 and 109.8 in 2024-25 — a point, against a median district’s whole three-year range of 2.1 points on the Index itself. 121 of 607 districts improved on the Index over that window and declined as a share of the maximum. verified computed One district in five went backwards on the measure the star rating is awarded on while going forwards on the measure the corpus quotes.

The count is exact and its edge is soft, so take the mechanism rather than the cut. 28 of the 425 districts that rose on the Index change share by less than a thousandth in either direction, and the nearest changes by five millionths. What does not depend on where the cut falls is the denominator itself: it rose 0.919 per cent, so every district’s share fell by that much relative to its own Index, and the 121 are the subset whose gain did not cover it. verified computed

This is a second and stronger reason for the rule Performance Index already states as “use the Index, not the star rating”. Coarseness was the first: a rating bins many districts together. Relativity is the second: a rating is scored against a denominator that other districts move. inference

The distributional constraint does not bind # ↑ Contents

Division (D)(4)(b)‘s requirement that “more than half of all districts or buildings do not earn the same performance rating” is satisfied on the achievement component with room to spare — the modal rating is four stars at 227 of 607, or 37.4%. verified computed It is a weak constraint, and the same column shows why: on that component three districts hold one star and 91 hold five. Nothing in the section requires the scale to be centered, only that it not collapse, and the corpus holds no year in which any component came close to collapsing.

What a rating has bought, and how often # ↑ Contents

Six budget acts each carried a property saying they were “not assessed against accountability-regime here”. Asked once instead of six times, the question is where an Ohio rating has determined a funding amount — and the twelve committed LSC greenbooks cover every biennium from FY2002 to FY2024.

Inside the greenbook window a rating has driven a formula payment in three episodes. The terms that occur only where one does — academic distress index, subsidy for high performance, graduation bonus, third grade reading bonus, performance bonus — appear in five of the twelve greenbooks and no others: H.B. 119, H.B. 153, H.B. 64, H.B. 49, and H.B. 166, whose freeze kept the bonuses running for community and STEM schools. 7 of the 12 bienniums carry none at all. verified crates/project, rating_payments::quiet, pinned by test

The census is of formula components, and the claim is no wider than that. Rating-keyed appropriations outside the formula did not end with H.B. 95: H.B. 110’s greenbook carries the Quality Community Schools payment, $54.0 million a year to community schools meeting standards that include report card grades, and an e-school pilot for dropout-recovery schools that exceeded standards on the graduation component. verified crates/project, rating_payments, pinned by test Neither goes to a district, and neither is in the table.

FY2004-05   H.B.  95   earmarks to academic emergency districts     on failure   outside the formula
FY2006-07   H.B.  66   EdChoice eligibility on building ratings     on failure   deducted from the district
FY2008-09   H.B. 119   Closing the Achievement Gap                  on failure   0.15% of the formula amount
FY2012-13   H.B. 153   subsidy for high performance                 on success   $17 a pupil, no state share index
FY2016-19   H.B.  64   graduation and third grade reading bonuses   on success   7.5%, times the state share index
FY2020-21   H.B. 166   the same bonuses, community and STEM only    on success   7.5%, no state share index
FY2026-27   H.B.  96   performance supplement                       on success   $13 x stars, no state share index

6 of the couplings in the table are formula components — the H.B. 64 row is two — and the first two rows reach a district’s money without passing through the formula at all. verified

The direction reversed in FY2012, and the equalizer was the exception. Closing the Achievement Gap paid more where more buildings were rated badly, and only to districts whose poverty index was 1.0 or above — thirty-one qualified. H.B. 153’s subsidy paid $17 a pupil to districts and community schools rated Excellent or Excellent with Distinction, flat at any valuation. The bonuses paid more where results were better, at fifty times the rate of the first episode, and multiplied by the state share index, which makes a payment larger in a poorer district. The supplement now in force pays on results and multiplies by no wealth term at all. So of the four payments on success, only the FY2016 district bonuses carried an equalizer. verified crates/project, rating_payments, pinned by test

The un-equalized form is the one that survived. H.B. 166’s freeze held every district and JVSD at its FY2019 foundation aid, so a district’s bonus stopped moving with its report card, while requiring community and STEM schools’ bonuses to be recalculated each year at a frozen formula amount. That version had never used the state share index — LSC calls it “identical to the calculation of the payment for traditional districts except that it does not use the state share index” — so for two years the only Ohio schools whose funding moved with their own results were charters and STEM schools, on the un-equalized calculation Ohio then readopted for everyone — and had first used in FY2012. verified

What that costs is measured at performance supplement rate: per pupil the present supplement runs 2.56 times higher in the least-poor fifth of districts than in the poorest. verified

Ohio built the measure long before it priced it. The report card predates the first formula coupling by years, and H.B. 59 — the act that rebuilt the card as A-F — coupled no new dollar to it. Its guarantee base held each district’s FY2013 high performance subsidy, which keeps a dollar an old rating bought without letting a new one move it. verified Whether the couplings were designed against the composition effect this node’s own findings describe, or arrived at independently of it, is not established. open