Career-Technical Planning District Membership

Source. Ohio Department of Education and Workforce, Ohio School Report Cards data API. reportcarddataapi.education.ohio.gov/api/GetByIrn/{irn}/{schYear}, plus the report card’s own organization index live-search.json. Type. Primary source — the department’s own roster of each planning district’s members. Location. An endpoint, not a file. This is the only source in the catalog that is.

Why it was retrieved. R.C. 3317.014(E)(1)(a) pays the lead district of each career-technical planning district $3 for every pupil in the summed enrolled ADM of that planning district’s members. Ohio Career-Technical Planning Districts gives the 92 leads and records membership as a live question — “a membership roster is a live question, not a settled absence. open”. This is the roster.

What it contains. For each planning district: its name, county, address, superintendent, and detail.enrollmentDist — one entry per member, with that member’s career-technical enrollment. For each district, the same endpoint returns ctpd and ctpdIrn, which state the planning district it belongs to directly.

A 954-row extract for FY2025 is committed at crates/dispersion/fixtures/ctpd-membership.csv, built by connect::fixtures::ctpd.

Why an API, when this catalog is otherwise files # ↑ Contents

Because it is the only route, and that is checked rather than assumed. The report card’s static download catalog publishes a Career Technical Planning District category of twenty-four files back to 2013 — CTPD-Ratings_2025.xlsx, CTPD-Grades_2018, and so on. Every one of them is ratings or grades. Membership appears in none. The two portals that did publish it are gone: webapp2.ode.state.oh.us/ctpd_region/ no longer resolves in DNS, and the report card’s own CTPD pages are an Angular shell returning one identical 11,166-byte document for every path, /robots.txt included.

It is nonetheless pinnable by URL, which is what makes it a source here rather than a procedure. The endpoint’s Azure Functions key — shipped to every visitor in the site’s main.<hash>.js — is accepted as a ?code= query parameter as well as a header, so each roster is a plain URL of the same shape as the report-card downloads already in the registry.

The key itself is not committed. The registry’s URLs carry {REPORT_CARD_KEY}, and edfund-connect fetch fills it from EDFUND_REPORT_CARD_KEY. Read it from the x-functions-key beside GetByIrn in the site’s current main.<hash>.js — the bundle’s name changes when the site is rebuilt, so find it from the page rather than guessing the filename. The committed fixture needs none of this; only a refresh does.

And the bytes are stable. Each document carries Cosmos metadata (_ts, _etag, _rid) and _ts is the last write, not the read: two fetches are byte-identical. So these pin by SHA-256 the way a published file does, and a changed digest means the department rewrote the record.

The 92, the 90, and the two # ↑ Contents

The CTPD table has 92 rows and its map legend “disagrees with itself” — the caveat Ohio Career-Technical Planning Districts records as unresolved. It resolves here. Ninety planning districts are rated and answer. The two that are not are the correctional institutions 200600 and 200602: they return HTTP 204 No Content and appear in no organization index, because nobody rates them. 90 + 2 = 92, and the 204 is a structural answer rather than a transient one.

Caveats # ↑ Contents

• enrlCtpd is career-technical enrollment, not enrolled ADM. The statute pays on enrolled ADM and this source does not publish it. Every share computed from this fixture describes who does career-technical education through a planning district, not the base the payment multiplies, and nothing built on it returns a dollar.

• A roster entry names its member and gives no identifier. enrollmentDist entries carry distName and enrlCtpd and nothing else. Resolution is against the report card’s own organization index — 4,018 organizations with IRN, kind and name — rather than a match against this repository’s panel.

• Nineteen names are shared by forty-seven districts, and no roster can tell them apart. Normalized for the two styles the publisher writes, 607 district names collapse to 579 keys: three Perry Locals, three Buckeye Locals, three Madison Locals. The index disambiguates with a - County tail the roster does not have. Those forty-seven are therefore retrieved individually, because a district’s own record carries ctpdIrn. No two same-named districts share a planning district — checked, not assumed — so the placement is unique, and the builder refuses a shared name it cannot place rather than picking a candidate.

• 347 of the 954 members are not districts. They are community schools under R.C. Chapter 3314 and STEM schools under Chapter 3326 — the population the 609-district panel does not carry. They are 36.4% of the membership and 4.75% of career-technical enrollment, and 277 of them carry none at all.

• The organization index is published per school year under the current year’s prefix (2026appsettings/). A prior year’s prefix is not guaranteed to survive, so the index is a current-roster artifact and a re-retrieval years hence may not reproduce this one.

• The key and the bundle filename both move when the site is rebuilt. If the department rotates the key, every one of these sources returns 401 at once — a loud failure rather than a quiet one, and the repair is to re-read it from the current bundle. No amount of care here prevents the rotation; committing the key would only have made the failure later and more confusing.

• FY2026 answers on the same endpoint and is not committed. FY2025 is taken because it is the year whose membership was cross-checked against the other side of the same API: every district’s own record carries a ctpdIrn, and all 607 agreed with the roster-derived placement.