Data Changelog

Real, dated changes to the data and methodology behind PlainTeacher - not a list of every site update, only ones that affect what a figure means or how it was calculated. See our corrections process for how a reported issue reaches this list.

- Upgraded to FY2023 public NCES files; source and trend series refreshed

PlainTeacher now uses NCES’s public FY2023 (2022-23) F-33 provisional v1a file and the matching 2022-23 CCD Staff Universe file. The loader selects the unique NCES Teachers derived-major-category FTE row for each district, preventing other staffing categories from changing a teacher-salary denominator. The database now contains 13,149 districts, with 12,156 reported-FTE calculations and 57 documented enrollment-based fallbacks. The historic F-33 panel now runs FY2019-FY2023 on one estimated basis, and the state purchasing-power release now joins the same-year 2023 BEA Regional Price Parities. Source freshness is checked from declared direct public downloads before each rebuild. FY2023 remains a provisional NCES release; see methodology for limits.

- Salary numerator corrected to teacher-only pay; every figure restated

Average salaries were calculated from the F-33 survey's total instructional salary line (variable Z33) divided by a teacher headcount. That line also pays instructional aides, paraprofessionals and substitutes, so charging it to teachers alone overstated pay, nationally by about 27% against NEA's published average for the same year. All salary figures now use the survey's teacher salary items (Z35-Z38), which NCES defines as base salaries paid to certified teachers. The national average falls from $73,045 to $68,021 against NEA's reported $66,745, and state and district figures were restated accordingly. State averages are also now a plain mean across districts, matching how the site has always described them, instead of being weighted by student enrollment. Three states, Alaska, Illinois and Utah, file no teacher-salary breakout at all; their pages now show the spending and enrollment data they do report and no salary figure, and they are excluded from salary rankings, which run 1-48. See methodology.

- Salary estimate methodology corrected

Districts with a reported teacher headcount (NCES CCD Staff Universe Survey) now use that real figure instead of an enrollment-derived 15:1-ratio estimate. The enrollment-ratio estimate is now used only as a documented fallback for districts with no reported headcount (see methodology).

- Absurd estimates removed for non-enrolling agencies

Roughly 967 specialized service agencies (county offices, regional cooperatives) were showing an inflated per-pupil/salary estimate because the flat enrollment-ratio formula does not fit their staffing model. These are now flagged as unreliable rather than displaying a misleading number.

- Per-pupil spending artifact corrected

Fixed a denominator artifact where a specialized agency's small directly-enrolled headcount produced an implausibly large per-pupil spending figure. These agencies now show total spending and enrollment instead of a per-pupil rate.

- Methodology re-verified against live NCES documentation

Confirmed the FY2022 F-33 vintage cited throughout the site is still the current Final release. A newer FY2023 dataset exists upstream but remains provisional (not yet Final) as of this check; we will not migrate to it until NCES publishes the Final version, to avoid re-doing the work on data that could still change.

- Site launched

Initial data load: NCES CCD F-33 Local Education Agency Finance Survey and Staff Universe Survey, FY2022 (2021-22 school year).

Don't see something you'd expect here? Not every internal fix is data-facing - this list covers changes that affect what a figure means or how it's calculated, not site design or performance work. If you believe a displayed figure is wrong, let us know.

PlainTeacher changelog entries document real pipeline and methodology changes against NCES CCD F-33, no number is typed in by an editor. Only data-facing corrections appear here. Report a figure that looks wrong via /contact. See our editorial standards & corrections policy, the methodology behind these numbers, or report a data error on this page. Data current as of August 2026.