Census block geography: school districts, legislative districts, and population

Source. U.S. Census Bureau, Geography Division and Redistricting Data Office. Three files: BlockAssign_ST39_OH.zip (2020 Block Assignment File), sldl24.zip and sldu24.zip (2024 State Legislative District Block Equivalency Files, lower and upper chamber), oh2020.pl.zip (2020 Census Redistricting Data, P.L. 94-171). Type. Primary source — the Bureau’s own record of which geographies contain which of Ohio’s 276,428 census blocks, and how many people live in each. Location. www2.census.gov/geo/docs/maps-data/data/baf2020/, www2.census.gov/programs-surveys/decennial/rdo/mapping-files/2025/2024-state-legislative-bef/, and www2.census.gov/programs-surveys/decennial/2020/data/01-Redistricting_File--PL_94-171/Ohio/.

What they contain. The block assignment file maps each Ohio census block to the geographies containing it — among them SDUNI, the unified school district. The block equivalency files map the same blocks to the state legislative districts of the 2024 plan, for both chambers.

Ohio’s constitution requires each Senate district to be exactly three whole House districts, and the block data confirms it with no exceptions. So the Senate crosswalk could have been produced by grouping the House one and would have been exact rather than approximate. It is read from the Bureau’s own upper-chamber file anyway: the composition is not sequential — Senate 2 is House 44, 75 and 89 — so a derivation would need the mapping regardless, and a rule worth relying on is worth checking against the file that would break it.

The redistricting file carries each block’s population and its population aged 18 and over; the difference is the under-18 count, 2,591,886 for Ohio, 22.0% of the state.

The Senate view is also substantially less of an estimate. Seats three times larger mean 392 of 609 school districts sit wholly inside one, against 270 for the House.

Why this is here. Ohio’s school funding system contains no legislative district. The department computes funding per school district and stops, so there is no published mapping from school districts to House districts and no published funding figure for one — not because it is secret but because nobody in the system has ever needed the number.

Building it needs block-level geography because school districts and House districts do not nest. 339 of Ohio’s 609 school districts have population in two or more of the 99 House districts; Columbus has population in eleven. County can be an attribution — the department assigns each district exactly one, and the corpus inherits it. A House district cannot.

The trap in the first file, which is worth stating plainly. BlockAssign_ST39_OH.zip contains its own BlockAssign_ST39_OH_SLDL.txt, sitting in the same archive as the school district file and looking exactly like the right input. It is the 2020-cycle map. Ohio redistricted afterwards, and 66.3% of Ohio’s census blocks are in a different House district under the plan now in use. A crosswalk built from the convenient file would have been wrong for two-thirds of the state and would have looked entirely plausible — 99 districts, every school district placed, totals reconciling. Hence the separate 2024 source, which exists because Ohio is one of eight states with changes to both chambers in that cycle.

What the corpus does with them. Joins all three on the block identifier, aggregates to (chamber, school district, seat) triples, and expresses each as a share of the school district’s under-18 population. That is crates/project/fixtures/legislative-district-crosswalk.csv, 1,947 rows — 1,085 House and 862 Senate — covering all 609 districts in the funding panel with no losses in either chamber.

The join to Ohio’s own identifiers runs through the NCES Common Core of Data directory, whose ST_LEAID column is the IRN.

What it does not support. Anything stated as a fact about a House district’s school funding. Every apportioned figure is an estimate: the weight is under-18 population, which is a proxy for pupils and counts children in community schools, private schools and none; and it is a 2020 count applied to a FY2027 model, so a district that has grown or emptied since is weighted as it was. The one exact property is conservation — each district’s shares sum to one, so the apportionment loses no dollar. It is exact in the arithmetic and not in the published figures: seat totals are rounded before they are written, so on the committed feed the two chambers differ by $0.96, the House sitting $2.06 above statewide.realized_aid_total and the Senate $1.10.

Vintage risk the header assertion cannot catch. The crosswalk is pinned to the 2024 map. When Ohio redistricts again the file at the same URL changes and the fixture becomes silently stale: the column names do not move, so project::legislative_district’s header check will pass. The digest manifest is what detects it.