How we research, classify, and update this site

Definitions

We classify each state’s program using the following types. Where a program does not fit neatly, we note the ambiguity on the state page.

• Universal ESA — A state-funded, family-controlled education savings account open to all K–12 students, regardless of income or prior enrollment status.
• Limited ESA — An ESA with eligibility restricted by income, disability, district-of-residence, or other criteria.
• Voucher — A state payment that flows directly from the state to a participating private school on the student’s behalf.
• Tax-Credit Scholarship — A scholarship funded by donor contributions to a state-approved scholarship granting organization (SGO), with donors receiving a state tax credit.
• Refundable Tax Credit — A personal income tax credit for documented education expenses, refundable if it exceeds the family’s tax liability.
• None — No statewide private-choice program is currently in operation.

“Last verified” means the date on which an editor confirmed the key figures on a state page — award amount, eligibility, administrator, and application window — against the state program’s official published materials. It does not mean every sentence was re-written; it means the data was confirmed still accurate as of that date.

States without a program

States classified as “None” are verified annually to confirm that no statewide private-choice program has been enacted. We check the state legislature’s session records and the state education agency’s published program list. A “None” classification does not preclude local public-school choice, charter school enrollment, homeschool options, or federal programs — it reflects only the absence of a statewide private-school-choice funding mechanism.

Source hierarchy

For every state page, we work from the primary sources in this order: the enacted bill text, the codified statute, the most recent administrative rule, and the program administrator’s published guidance. Secondary sources — advocacy materials, news coverage, and educational podcasts — are used to surface changes and to triangulate when the primary sources are silent or ambiguous, but we do not cite them as authority.

Classification

We classify each state’s program into one of six categories: Universal ESA, Limited ESA, Voucher, Tax-Credit Scholarship, Refundable Tax Credit, or None. The boundaries between categories are not always sharp; we erred toward the design type that most closely matches the funding mechanism rather than the political branding.

Award amounts

We report the headline per-student amount as published by the administrator. Where amounts vary by grade level, income tier, or disability status, we report the most common award and note the variance in the state page narrative. We do not estimate or project; if the administrator has not published a current-year amount, we report the most recent published amount with the year.

Update cadence

State pages are reviewed on a rolling cadence, with priority given to states with active legislative sessions or imminent application windows. We re-verify award amounts and eligibility rules at the start of each application cycle. Major policy changes (a newly enacted program, a court decision affecting eligibility, an appropriation cap change) trigger an immediate re-publish.

Corrections

We publish corrections openly. When a fact on this site has been corrected, the page footer is updated with the date of the correction. Substantive corrections that change the editorial meaning of a piece are flagged inline.

What we will not claim

We will not claim that any state’s program is “the best” or “the worst.” We will not extrapolate from a single study to a general conclusion about whether school choice “works.” We will not advocate for or against the expansion of school choice. We will not infer parent or student motives from enrollment patterns.

The empirical research on ESAs and vouchers is mixed in ways that depend heavily on program design, geography, and which outcomes are being measured. We try to reflect that honestly, which sometimes makes our coverage less satisfying than the certainty offered by either side. That trade-off is intentional.

Page last reviewed May 25, 2026.