Diesel price connector
A decision record · diesel-price-connector
A twenty-second connector, eia-diesel, retrieving the Midwest diesel price — the first source here that is not about schools, admitted because R.C. 3317.0212 makes a fuel price an input to the funding formula whether or not the corpus holds one.
Context
Every connector approved so far retrieves something a school, a district, a court or the legislature produced. This one retrieves a commodity price, and the ordinary answer to a proposal like that is no: the corpus is about how Ohio funds its schools, and an energy series is somebody else’s domain.
The reason it is admitted is that R.C. 3317.0212 puts it inside the formula.
Divisions (C) and (D) name no dollar figure. Both of the state’s transportation rates — the $1,337.175 per weighted rider and the $6.867 per mile the FY2027 model carries — are defined as trimmed means of what districts themselves reported spending in the prior fiscal year, after excluding non-transporting districts and the ten highest and ten lowest per-pupil costs. Nobody chooses them. They are statistics over district behavior, and they move when that behavior moves, with no act required.
That makes the cost of running a bus a formula input, and diesel is the part of that cost with a published monthly price. Without it the corpus can say the rates are endogenous — the parameter node already does — and cannot say to what.
The gap this closes is a live one rather than a theoretical one. On a July-to-June fiscal year the Midwest series gives FY2025 $3.558, FY2026 $4.181, and FY2027 through August 2026 $5.121 — a 44% rise across two years. The FY2027 rates were computed from FY2026 reported costs, and the department’s calculator is dated 16 December 2025, before most of that move. The lag the section specifies is currently running about a fifth of a fuel price wide.
The decision
Approved. A twenty-second connector, eia-diesel, publisher U.S. Energy Information
Administration, feeding metric and fiscal-period, status Parsed rather than Wired: it
retrieves and parses to a committed fixture, and nothing downstream consumes it yet. It becomes
Wired when a calculator reads it.
One source: EMD_EPD2D_PTE_R20_DPG, No. 2 diesel retail prices for the Midwest (PADD 2),
monthly, from the hist_xls path. Cataloged at
Midwest No. 2 diesel retail prices — the EIA monthly series.
It is a separate connector rather than a second source under bls-cpi — the shape that was
considered first, since both are federal statistical series read for context rather than for
a funding figure. Rejected: a price index and a commodity price do different jobs here. The
CPI restates dollars so that two years can be compared; this explains why one of them moved.
Putting them behind one status flag would mean a diesel outage could report as a deflator
problem.
Three constraints are approved with it.
The fixture stays monthly. Every question this series will be asked is a fiscal-year one
and the temptation is to commit twelve averaged numbers. connect parses and does not compute,
and the aggregation is a real choice rather than a formality: fuel is bought across the year,
so a July-to-June mean is the natural window, where the deflator takes a June point
observation because an index level is not consumed. A monthly fixture can be re-aggregated;
an annual one cannot be taken apart.
The retail-price caveat travels with every figure. This is a retail price including excise tax, and Ohio’s is 47 cents a gallon on diesel under R.C. 5735.05(E)(2), refundable to school districts. The level is therefore wrong for a district budget.
The direction of the error is the part worth stating, because this decision originally stated it backwards. Subtracting a fixed per-gallon wedge from both ends of a rise widens the proportional change: FY2025 to FY2026 is +17.5% at retail and +20.2% net of the tax. The retail series understates what districts experience, so an elasticity fitted to it without the adjustment is biased low rather than high. No Ohio school-bus bulk price has been located; the catalog entry keeps that open rather than substituting a national benchmark.
PADD 2 is not Ohio, and the entry says so. The Administration publishes no Ohio retail diesel series. The Midwest region is the smallest published geography containing the state.
A note against the obvious next step: this does not approve fitting a fuel elasticity to
the F-33 transportation series. crates/dispersion/tests/f33_student_transportation.rs
establishes that the two largest moves in that panel are the FY2021 pandemic collapse and its
FY2022 rebound, and diesel collapsed in calendar 2020 at the same moment Ohio’s buses stopped.
The confound is nearly collinear with the treatment. A regression over the full panel would
recover the closure of Ohio’s schools and report it as a fuel response. What to do about that
is a later decision, and it needs the pre-FY2027 rate series that nothing here yet holds.