Nearly 28 years after New York legalized charter schools, here's what the public record shows
What the Charter Schools Act of 1998 built: 356 operating charter schools, 189,888 students, an estimated $3.69B a year in district-to-charter tuition payments, and what the public filings do and do not settle.
New York's Charter Schools Act passed in the small hours of 18 December 1998, in a special session of the Legislature. Governor George Pataki had signaled he would not sign a pay raise for legislators unless charter schools moved with it. Both passed.
A charter school is a publicly funded school run by an independent board rather than by an elected school district. It is exempt from many of the rules districts follow. In exchange, its charter — its license to operate — comes up for renewal every few years, and can be revoked.
The law capped the state at 100 charter schools and, unusually, gave half the authorizing power to the trustees of the State University of New York rather than to the Board of Regents, which oversees every other public school in New York. That split was the compromise that moved the bill.
The cap moved twice
- 1998 — 100 schools, under the original act.
- 2007 — 200 schools, after the first cap had been reached.
- 2010 — 460 schools, raised to compete for federal Race to the Top funding.
What exists now
Twenty-seven years and nine months on, DistrictFacts counts 356 operating charter schools enrolling 189,888 students. The 1998 compromise is still visible in who signs the charters: SUNY, the body that got half the original authority as a negotiating concession, now authorizes about two in three.
NYSED's published fact sheet counts 352 operating charter schools serving more than 180,000 students in 2025-26. DistrictFacts counts 356 and 189,888. The difference is the filing year: this analysis uses the 2024-25 school year, the most recent for which school-level financial and enrollment filings exist, while NYSED's headline is 2025-26. Our collection holds 363 charter records in total; 7 of those closed in 2024 and 2025 and are excluded from every figure here, and they held 355 students at their final filing.
How this count is built
Who authorizes New York's charter schools
- SUNY trustees
- 225
- NYSED / Board of Regents
- 94
- NYC Dept. of Education
- 35
356 operating schools; authorizer unrecorded for 2. Closed schools excluded.
Authorizer records, compiled by DistrictFacts
What the districts pay in tuition
When a student enrolls in a charter school, the school district where that student lives pays the charter a per-pupil tuition set by the state. For 2025-26 the published rates run from about $13,000 to about $34,000 depending on the district; the statewide mean is $19,320. New York City pays a single citywide rate of $19,903 for every charter school inside it.
Multiplying each school's paying-district rate by its reported enrollment gives an estimated $3.69B a year in district-to-charter tuition payments, across 355 schools. Charter schools are tuition-free to families — this is money moving from one public body to another, not a bill any parent receives. It is an estimate we computed, not a figure any single filing states, and it is worth being precise about how it is built. The method is enrollment multiplied by the applicable resident-district charter tuition rate. It applies the 2025-26 published rates to the most recently reported enrollment, which is 2024-25 for 347 of the 355 schools and 2023-24 for the remaining 8. Applying next year's published rates instead moves the total to $3.85B, so the rate year is worth about 4.5% either way. It does not adjust for special-education payments, midyear enrollment changes, or the reconciliations districts and charters settle between themselves after the fact.
There is one useful sanity check on that estimate. For the 140 schools that file their own Form 990 rather than sitting under a network return, the same calculation produces $1.32B against $1.64B of revenue those schools actually reported — a ratio of 0.81. That is roughly what it should be: basic tuition is the largest line in a charter school's revenue but not the only one, with federal grants, philanthropy, food service and rental income alongside it. It is not proof: Form 990 revenue covers a fiscal year that need not align with the school year, and includes federal funding, grants, philanthropy and food service alongside tuition. A ratio materially above 1.0 would be a warning that the estimate or its period alignment needed closer examination.
Where the money goes
Charter schools file federal Form 990s — the annual return required of tax-exempt organizations. Those filings are public. Across 159 filings covering 241 operating schools: $2.62B in reported revenue, $2.55B in reported expense, and $4.73B in net assets, the accumulated value an organization holds after debts. Net assets exceed a single year's revenue by about 1.8 times. The filings do not, on their own, say whether that is prudent reserve, accumulated buildings, or anything else.
Twenty-six filings report an amount on the management-fee line of their return — money a school sends to a company or nonprofit that runs some or all of its operations — totalling $123.4M. That line is not a complete picture of management spending: Brooklyn Dreams and Riverton Street each report nothing on it while separately reporting $9.6M and $16.1M to National Heritage Academies for management services elsewhere in the same filing. The two figures overlap and should not be added together.
About 100 of these filings are network-level: one return covering every school an operator runs, attached to each of those schools' records. Adding the field up school by school counts a network once per school, and reports $18.77B — 7.2 times the real figure. Every total above counts each filing once, by its federal employer identification number.
Method note
Who pays, and who receives
Under New York's charter law it is the district where a student *lives* that pays their charter school, not the district where the school sits. Our records identify each school's host district from its BEDS code, and we assign that school's whole enrollment to it. For New York City that distinction barely matters — the city is a single paying district and nearly all its charter students are city residents — and the city is 81.2% of the total. Outside the city it matters a great deal: a charter in one district routinely enrolls students from its neighbors, and each of those students is billed to their own district at their own district's rate. So read the statewide total as a total, and read the district breakdown below as showing where the schools are, not a bill any named district received.
Estimated district-to-charter tuition, by the district each school sits in
- New York City
- 81.2%
- 22 other districts combined
- 11.7%
- Buffalo
- 4.1%
- Rochester
- 3.0%
Schools are grouped by their host district, from the BEDS code. The paying district is the student's district of residence, so outside New York City these are approximations, not district liabilities. New York City is a single paying district for charter purposes.
Estimated from NYSED published tuition rates and reported enrollment
With New York City set aside, the districts hosting the most charter enrollment become legible. Hempstead and Uniondale stand out — two districts hosting schools with an estimated $89.6M and $41.9M of tuition between three charter schools — though how much of that is billed to those districts rather than their neighbors is not something these records settle.
Where the charter schools are, outside New York City
- Buffalo
- $150,900,000
- Rochester
- $112,100,000
- Hempstead
- $89,600,000
- Albany
- $62,200,000
- Uniondale
- $41,900,000
- Syracuse
- $41,400,000
- Kenmore-Tonawanda
- $28,200,000
- Riverhead
- $23,300,000
Estimated tuition for the schools located in each district, not that district's actual charter bill: students from neighboring districts are billed to their home district. Treat this as a proxy for scale. 8 of the 24 districts outside New York City; own scale, not comparable by eye with the shares above.
Estimated from NYSED published tuition rates and reported enrollment
On the receiving end, the Form 990 filings name each organization's largest contractors. Across 305 reported payments to 234 distinct recipients totalling $250.8M, the largest single relationship is between National Heritage Academies and five schools it manages, which together reported paying it $59.1M for management services.
Largest reported recipients of charter school payments
- National Heritage Academies (management, 5 schools)
- $59,090,000
- Hamilton Stern Construction (construction)
- $15,520,000
- Democracy Prep Public Schools (management)
- $13,750,000
- Signature Construction Group (construction)
- $8,920,000
- Excellence Community Schools (management)
- $8,390,000
- Prospect Schools (management)
- $8,270,000
- Explore Schools (academic and business services)
- $4,870,000
Amounts the paying organizations reported themselves, deduplicated by the payer's federal employer identification number. A Form 990 lists only an organization's largest contractors, so this is neither total spending nor a complete ranking of vendors. The filings do not say whether any price was reasonable or what was delivered for it.
IRS Form 990 filings, fiscal year 2023
Every organization named above was confirmed to exist independently of the filings, and each paying relationship was corroborated from the recipient's own public record: National Heritage Academies names Brooklyn Dreams, Brooklyn Excelsior, Brooklyn Scholars and Riverton Street among the schools it manages; Signature Construction Group lists Academic Leadership Charter School among its completed projects; Hamilton Stern Construction is a Rochester-area commercial builder and the paying school is in Rochester. One further payment — $4.07M reported as rent by Legacy College Preparatory Charter School — is excluded, because the recipient could not be identified in any public record beyond the filing itself, and we do not name an organization on the strength of a single document.
How these recipients were checked
What you can find out depends on who authorized the school
The 1998 law split authorizing power, and 28 years later that split decides how much of a school's oversight record is public. SUNY posts a dated table of its own oversight documents. The Board of Regents publishes its actions as narrative prose. New York City's Department of Education publishes per-school PDFs and no structured outcome list at all. Coverage of renewal outcomes tracks that difference almost exactly.
Renewal outcomes on the public record
- SUNY (180 of 225)
- 80.0%
- NYC Dept. of Education (19 of 35)
- 54.0%
- NYSED / Board of Regents (32 of 94)
- 34.0%
A difference in publishing practice, not in how closely any authorizer supervises. A school with no outcome on file has not necessarily gone unreviewed.
SUNY Charter Schools Institute, NYSED, NYC DOE
How students do
For 308 charter schools, state records let us compare a school's own test results against the district its students come from — its sending district. Charter schools' English proficiency runs a median 12 percentage points above their sending district, and math a median 14.3 points above.
Charter proficiency above the sending district
- Math proficiency
- 14.3%
- English proficiency
- 12.0%
Median difference in percentage points; 299 and 308 paired schools.
NYSED school report cards
The same records show the two groups are not serving identical populations. Charter schools enroll fewer students with disabilities and fewer English language learners than the districts around them.
Charter enrollment below the sending district
- English language learners
- 7.9%
- Students with disabilities
- 5.8%
Median shortfall in percentage points; 339 and 357 paired schools. Both charts come from the same NYSED report cards, and the record does not say how much of one explains the other.
NYSED school report cards
What the record does not settle
The public filings answer what was spent, who received it, who authorized a school, and how its students scored. They do not answer whether a management fee was fair, why a network holds the reserves it holds, or what would have happened to these students in a district school.
It is worth being blunt about the limits, because these figures will be quoted by people arguing in both directions. What follows does not follow from this data:
- That charter schools caused district enrollment to decline. The records show both happening; they do not show one causing the other.
- That the $3.69B would otherwise have remained available for identical district spending. Funding formulas, fixed costs and enrollment-driven aid all move when students move.
- That charter schools are more or less efficient than district schools. Different populations, different filings, different cost structures.
- That charter expansion produced any particular academic result. The proficiency gap and the composition gap sit in the same records, unseparated.
- That families are better or worse off because of the 1998 law. No filing answers that question.
Explore the records yourself
Every figure above sits on a page you can open. Three questions worth asking of the charter schools near you:
- Which body authorized them, and when does the charter come up for renewal? Every charter page names its authorizer and term.
- Does the school report a management fee, and to whom? The finance section carries the Form 990 figures and links to the filing.
- How does its share of students with disabilities and English language learners compare with the district around it? Both sit side by side on the school's page.
DistrictFacts Data Desk
- Data period — 2024-25 filings, with 2025-26 published tuition rates.
- Analysis generated — 8 September 2026.
- Records analyzed — 356 operating charter schools (363 held; 7 closed and excluded).
- Sources — IRS Form 990; NYSED State Aid charter basic tuition; NYSED school report cards; SUNY Charter Schools Institute, NYSED and NYC DOE for authorizer status.
- Reproducible — every figure regenerates by running scripts/charter-act-facts.ts.
- Corrections — request one from this page; corrections appear at the foot of the story and in the public corrections log.