District Profile Report — the "Cupp Report"

Source. Ohio Department of Education and Workforce, District Profile Reports, published annually. Named after Bob Cupp — the same legislator who later co-authored the Fair School Funding Plan. Type. Primary source — machine-readable per-district finance data. Location. education.ohio.gov, Finance and Funding → School Payment Reports → District Profile Reports. The connected FY2024 edition is FY24-District-Profile-Report-Revised-12-18-2025.xlsx, about 1 MB — the department’s revision of the original ...-Final-12-12-2024.xlsx, adopted by #439. The caveat below says what moved.

What it contains. Sixty variables for each of Ohio’s 606 traditional school districts, on five worksheets — District Data, Similar District Data, Statewide Data, and a formatted one-page-per-district report. Demographics, personnel, property valuation and tax data, local tax effort, expenditure, and revenue by source, each with comparison columns against similar districts and the state.

This is the single most useful source the corpus has found. It carries, per district and in one file: enrolled ADM, economically disadvantaged share, classroom teachers’ average salary, assessed valuation per pupil, current operating millage, effective Class 1 and Class 2 millage, operating expenditure per pupil, and revenue split by state, local, federal, and other. It is what turned the exemplar agencies from identifiers into series and what let the corpus answer how many districts sit at the 20-mill floor.

A seventeen-column extract is committed at crates/dispersion/fixtures/cupp-fy24-district-data.csv and is the basis of that crate’s integration tests: the ten columns the equity findings read, and since #408 the report’s seven personnel columns. edfund-connect rebuild writes it from the cached workbook through connect::fixtures::build_profile_extract.

What the personnel family is, and is not. Columns O through U of District Data: classroom teachers’ average salary; the share of teachers with 0-4, 4-10 and 10+ years’ experience; FTE Number of Administrators; administrators’ average salary; and the pupil-administrator ratio, which is the department’s own quotient of enrolled ADM by the FTE count. One of the seven is a count of staff, and it is administrators. There is no teacher FTE and no position category finer than that. The report’s page defines the count as the sum of twelve EMIS position codes — administrative assistant (101, the licensed code), deputy superintendent (103), assistant principal (104), principal (108), superintendent (109), supervisor/manager (110), treasurer (112), coordinator (113), education administration specialist (114), director (115), community school administrator (116) and other officials/administrators (199). That is the union of R.C. 3317.011’s four administrator elements at (F)(1)-(3) and (G)(1), so the count can be set against the two-other-administrators floor and against nothing else in that section; the six other staffing floors meet no column here, and the category detail they would need lives in EMIS staff reporting, which is not cataloged.

Access constraints. Freely available. XLSX rather than CSV; it parses with stdlib zipfile plus xml.etree — an XLSX is a zip of XML — with no third-party library required. The District Data worksheet is xl/worksheets/sheet2.xml.

Caveats, all observed rather than assumed:

• Fiscal and tax years are mixed within a single row. Enrolled ADM is FY2024, assessed valuation per pupil is FY2023, millage is TY2023, and OFCC adjusted valuation is FY2025. The column headers state this; a naive read that treats a row as one year is wrong.

• Missing values appear as #N/A and #DIV/0!, not as blanks. Southern Local (046441), Columbiana County reports no operating expenditure per pupil at all, so statewide expenditure statistics are computed over 605 districts, not 606.

• Published rounding produces impossible-looking rows. Five districts report an effective Class 1 rate marginally above their voted operating millage — up to 0.0004 mills, e.g. 22.65 effective against 22.6499 voted. These are artifacts, not violations, and a validity check with too tight a tolerance will reject good data.

• Traditional districts only. No community schools, no STEM schools, and no JVSDs, which is why the corpus’s JVSD exemplar cannot be populated from here.

• The department revised the FY2024 edition a year after publishing it, and the revision is the connected file. The original FY24-District-Profile-Report-Final-12-12-2024.xlsx (workbook modified 2024-12-27) was connected until #439; the page now serves FY24-District-Profile-Report-Revised-12-18-2025.xlsx (modified 2025-12-18), and the original URL still resolves and still serves its pinned bytes — re-fetched 2026-09-23, digest unchanged. So the original was not kept out of necessity and is not lost by moving.

  Why the revision was adopted. It is not a reissue. The department corrected a genuine defect on the aggregate sheets: every figure on Statewide Data changed from an unweighted mean of district values to an enrolled-ADM-weighted mean, reproduced exactly here on eight of the columns. The original’s “statewide” Black enrollment share was 10.5%, which is the average of 606 district shares; Ohio’s share is the revision’s 17.3%. Its “statewide” FTE Number of Administrators was 21.35, which is the mean district’s count and not a state total. A source whose publisher has withdrawn its own aggregates for cause is the wrong file to keep as the corpus’s statement of FY2024, even where the corpus does not read those cells. The corpus does not read Statewide Data at all — rebuild reads District Data only, and computes its own statewide statistics from district rows. Which convention those use is a separate question, filed as #463.

  What moved in what the corpus reads. Compared cell by cell, joined on IRN — the revision also re-sorts District Data from district name to IRN, so a positional comparison is meaningless. At the extract’s own precision the answer is three columns: the shares of teachers with 0-4, 4-10 and 10+ years’ experience, for 293, 292 and 293 districts. The teachers’ average salary (538 districts) and the administrators’ average salary (439) differ only below the fourth decimal — 68412.482997679996 against 68412.482997689003 — and do not reach the fixture at all, so the earlier reading that a re-vendor moves “four salary and experience columns” overstated it by two. FTE Number of Administrators is identical in both, as are the ten columns the equity findings read. No crate reads the three experience shares, no figure binds one, and the full crate suite is green on the rebuilt fixtures.

  The 25 doubled-space names cost nothing in the end. The revision restates 25 district names as Orange City (046581). clean_name now collapses internal whitespace, which it had to anyway — substituting a space for a comma in EDGE ACADEMY, THE was manufacturing the same defect in five other fixtures — so the committed names are unchanged and still match the calculator’s spelling.

• The extract’s row order is this repository’s, not the publisher’s. build_profile_extract sorts on IRN since #439. It wrote rows in sheet order until then, so the revision’s re-sort moved all 606 rows and the three that changed could not be seen in the diff. The one-time churn of adopting the sort is in that commit; a future re-sort upstream is now a no-op.

Also read by #

• FSFP Base Cost Calculation — the administrator count against the section’s funded administrators

Feeds connector #

dew-foundation, and via tax-abstract for the millage and valuation columns.