Ohio Educational Directory System: nonpublic school extract

Source. Ohio Department of Education and Workforce, Ohio Educational Directory System (OEDS), the public data extract. Type. Primary source — the department’s own register of organizations, generated on request. Location. https://oeds.education.ohio.gov/DataExtract/GetRequestOrgExtract, as a POST of one form field, jsonData, naming the organization types wanted. Type 5 is Nonpublic School. A GET of the same URL is refused with a 403 by the department’s gateway, so the registry carries the form (connect::registry::REQUESTS) rather than only a URL.

What it contains. One row per open organization per school type, 24 columns: building IRN, name, school type, grade span, status, designated county, web address, email, mailing address, phone, principal, and a parent IRN and name. For nonpublic schools the parent is the diocese or association — the Cleveland Catholic Diocese is parent to 114 — and never the public school district. There is no district column. Every IRN is written as an Excel formula, ="057539", so a spreadsheet keeps the leading zero, and the first line is Generated on <timestamp> ahead of the header.

When first fetched it listed 787 open nonpublic schools in 903 rows: a school offering elementary and middle grades appears once per type, with the same name and address. The extraction collapses them and fails if two rows for one IRN ever disagree.

Why this is here. R.C. 3317.024(E)(1) pays auxiliary services through the district a chartered nonpublic school is located in, and no October enrollment file says which district that is (Chartered nonpublic enrollment by building, 1977–2025). This extract is the only public record of where a nonpublic school is: an address against a building IRN, which the October files share. The address is geocoded into the 2020 census blocks by the Census Bureau’s batch geocoder, and the block assignment file gives the district each block lay in. crates/project/fixtures/nonpublic-school-district.csv is the result. The decision record nonpublic-school-location sets out why that is a keyed join with a stated reach, not the geographic match nonpublic-enrollment-connector rejected.

What it does not support.

• Anything before the day it was generated. It is a snapshot: schools open that day, at that day’s address. A school that has moved is placed where it is now, and a school that has closed is absent. Six buildings in the October 2025 count are already missing from it, and 23 of October 2023’s.
• A digest that means anything about the schools. The Generated on line changes on every fetch, so the digest moves whether or not a school did.
• The Selected field. The request carries a list of columns wanted and the extract ignores it, returning all 24 whatever is asked for.
• Closed organizations. The directory’s search API (Api/searchOrg) reaches them and is not read.