NH / FOLLOW THE EVIDENCE

School funding.
Facts need context.

What the numbers show. What the ballot changes. And how New Hampshire got here.

A fair, independent check of NH School Facts, with the history behind 2026 Question 2.

Updated October 24, 2026 • Statewide scope
Original agency records, enacted laws and court decisions

Start with the historyUnderstand Question 2Compare the claimsOur assessment

How we got here

The court cases ask what the state must fund. The ballot asks how local revenue and central-office budgets may be limited. Those are related debates—not the same legal question.

  1. 1. A statewide duty grew out of an unequal property-tax system.

    The 1993 and 1997 school-funding decisions established the state’s duty to provide a funded opportunity for adequate education and requirements for proportionate taxation. Communities have different property-tax capacity; the same tax effort does not produce the same resources per pupil. The 2025 ConVal opinion restates those holdings.[11]

  2. 2. Legislators built—and repeatedly changed—the financing rules.

    The 1999 law created a uniform education property tax alongside other revenues. The 2005 law fixed the SWEPT target at $363m. Later decisions changed retirement support and how excess SWEPT is retained. These choices affect the burden left locally; they do not explain every spending increase.[14] [15] [19] [12]

  3. 3. In 2025, ConVal found inadequate state base funding.

    The Supreme Court affirmed that the challenged base-aid statute was unconstitutional and upheld a conservative minimum guidepost above $7,356.01 under the existing framework. It reversed the immediate-payment directive. The case did not declare school expense increases unconstitutional, and the guidepost is not the measured full cost of every school service.[11]

  4. 4. In 2026, elected officials enacted a separate local cap vote.

    HB1300 passed its final conference votes 183–170 in the House and 15–8 in the Senate. Governor Kelly Ayotte signed it July 15. It places the cap question before voters in 2026 and 2028. The governor’s stated case is taxpayer protection. The law limits local levy growth and central-office budgets; it does not itself appropriate a ConVal funding remedy.[56] [58] [55]

  5. 5. The decision now involves both costs and consequences.

    The choice is not resolved by a single high-spending statistic. Restraining a levy can restrain tax growth; whether it reduces avoidable overhead or constrains needed services depends on costs, other revenue, the statutory definitions and overrides. That is the budget logic to test—not a claim that every district will have the same result.[55]

This sequence documents laws, holdings and their financing relationship. It does not establish the motives of NH School Facts or prove that the site was created to overturn a particular judgment. Read the full dated policy timeline →

What Question 2 actually does

HB1300, Chapter 324, creates a local statutory adoption question. It is not a statewide vote to reverse ConVal or a constitutional amendment. The enacted law—not a share graphic—is the starting point.[55]

RuleWhat it limitsWhat it does not guarantee
Local school tax levyPrior-year non-bond levy × (1 + inflation + net new taxable property growth). General market appreciation and reassessment are excluded from the growth term.[55] Unchanged classroom services, an individual tax bill falling, or the state making up a revenue gap.
SAU central-office budget6% of the member districts’ total appropriations, excluding certified bonded capital. The law defines the functions and excludes school-based administration and listed services.[55] That all Census “administration” belongs in the numerator, or that the separate district levy cap cannot affect services.

Adoption, timing and exceptions: the law requires three-fifths approval, with separate district and SAU tallies. Overrides are provided, with procedures depending on governance form. A November 2026 approval begins with FY2028; a November 2028 approval begins with FY2030. Certified bonded-capital costs are excluded; the scheme expires January 1, 2032.[55]

The central framing problemClassroom instruction is excluded from the central-office category. That is not the same as being insulated from the consequences of a cap on district tax revenue. Conversely, a cap does not prove classroom cuts will occur everywhere. Both categorical assurances exceed the budget evidence.[55]
Read the exact ballot wording in the enacted law
“Shall the [name of municipality] limit property tax growth for [name(s) of school district(s)] under RSA 32:5-i? If adopted for a two-year period: (1) the local property tax levy may not grow beyond the prior year’s amount, adjusted for inflation and new construction; (2) SAU central office spending may not exceed 6 percent of total school district appropriations; and (3) bonded capital costs are excluded from both limits. These caps apply only to administrative operations of the SAU central office and do not affect classroom instruction, school-based services, or other municipal expenditures. These limits may be overridden as provided in RSA 32:5-i. Adoption requires a three-fifths (3/5) majority vote.”

RSA 32:5-i V, as enacted by Chapter 324. Placeholders above are in the law. The operative definitions and procedures continue in VI–XVII.[55]

What the September court order says: the Supreme Court denied a petition for original jurisdiction, citing Rule 11. Its one-page order gives no substantive explanation. It should not be described as a judicial finding that the cap cannot affect classrooms.[57]

Compare at a glance

Some figures are accurate. Some conclusions change when the year, denominator, cost category or legal question changes. Tap a claim for its individual check.

Claim or framingIndependent checkAssessment
Real pay +8%; spending +45%Both reproduce; the spending definition changes[76] Arithmetic is not comparability
More state aid does not lower taxesAid and taxes rising together do not identify aid’s effect[65] Causal claim unsupported
73% more spending, four points higher73% / four points and 33% / 26% disadvantage both reproduce[71] Interpretation needs limits
Only administration is affectedTwo caps: local levy + central office[55] Misframed
Administration costs $2,140 per pupilCorrect; includes school administration[1] Supported number
101 SAUs; 50 under 1,000 pupilsBoth counts reproduced; $62m savings still an old model[48] Counts supported
Real cost per pupil rose 82%The rounded agency endpoints support it[6] Supported trend
EFAs are about a penny per dollar1.24% mixes years; aligned-year projection ≈0.73%[22] Mixed basis
EFA award below school average = savingsAverage cost is not the cost actually avoided[41] Inference unsupported
NH’s combined NAEP loss was three times U.S.Rounded sum checks; it is not an official NAEP index[36] Arithmetic supported
The ballot overturns the funding rulingState funding duty and local cap are separate[11] Incorrect linkage

How truthful, transparent and balanced is the original?

Its raw figures are substantially stronger than the conclusions it draws from them.

This is an editorial assessment of the statewide main and share pages reviewed October 24, 2026. The scores summarize the checks below; they are not a statistical truth rate or a finding about the authors’ private intentions.

DimensionScore / 10Why
Factual accuracy of reported figures8 / 10Many prominent figures independently reproduce: administration, staffing ratios, enrollment, recent financing, teacher-pay growth and the district comparison. Some comparisons mix years or measures, and the NAEP sum is a constructed metric. The source trail often needs reconstruction to identify the right release and denominator. Pay · Financing · Scores
Honesty of presentation: transparency and context5 / 10Source links and several real caveats deserve credit. The district section discloses equal weighting, small districts and higher disadvantage. But categorical “fact” conclusions go beyond descriptive comparisons; the limits are less prominent than the headlines. This evaluates what a reader is told, not whether an author knowingly deceived anyone. EFA costs · Funding duty
Balance / freedom from advocacy framing3 / 10The selection and presentation consistently favor a spending-and-staffing explanation and school choice. The site does show the falling real state-funding contribution, which complicates its own frame. But competing explanations and the constitutional funding distinction receive much less attention than the preferred conclusion. State versus local · Legal history
How to use itTreat it as a substantially sourced advocacy presentation. Check the inference as carefully as the number. Accurate data can establish that costs rose while scores fell; that does not establish what scores would have been with a different funding policy.
How these judgments were made

Each dimension uses the same broad anchors: 0 means fundamentally unreliable on that dimension; 5 means mixed quality with material weaknesses; 10 means consistently strong. Accuracy asks whether inputs, units, dates and arithmetic hold up. Transparency asks whether source identity, uncertainty and material caveats are visible. Balance asks whether the page fairly confronts evidence that could weaken its favored explanation. These are reasoned, subjective ratings, not a claim that exactly 80% of sentences are true.

No penalty is assigned merely because a source advocates a policy or a figure was initially difficult to retrieve. The audit now credits figures recovered from original releases. Numerical agreement does not automatically validate a causal claim. A conclusion is also not false merely because it favors lower spending or school choice; the problem is presenting an untested inference as settled by the displayed data.

Question 2’s legal framing is covered separately as requested. The rating does not charge NH School Facts with ballot wording it did not write, or assert that the site was created to overturn a particular court judgment. Neither this audit nor its score establishes the authors’ motives.[79] [80] [11]

Each claim, checked

Every point keeps its own comparison, precise interpretation and source trail. Claim wording is paraphrased; exact numerical anchors and short excerpts identify the original. Ballot-context checks are clearly identified.

A / Primary evidenceOriginal agency record, enacted law or court-authored decision inspected.
B / Reproducible calculationInputs, units and denominator visible. The formula does not itself prove a cause.
C / Not independently reproducedThe exact attributed figure or original method could not be verified. A documented alternative is shown where available; this does not establish falsity.

Print includes every claim and its detailed evidence, regardless of active screen filters.

01BallotThe “only administration” framing is misleading

Question 2 contains two separate limits.

The 6% rule targets central-office budgets. The separate rule caps the district’s local tax levy. That second limit is broader than administrative costs and can constrain the revenue available for classroom services.[55] [56]

Ballot context

Two-year levy limit + 6% central-office limit

Context: the ballot wording says the caps apply only to SAU administrative operations and do not affect instruction or school-based services.

Read the enacted law ↗
What the primary record shows

Chapter 324 inserts RSA 32:5-i. Paragraph VIII limits prior-year levy growth, excluding certified bonded capital, to inflation plus net new taxable property growth. Paragraph IX separately defines the 6% central-office cap and its exclusions. Adoption requires three-fifths, with separate district/SAU tallies; overrides are provided by paragraph X.[55] [56]

What this means, in precise terms

Question 2 is not limited to administrators. It pairs a central-office cap with a separate cap on the district’s local tax levy. Excluding classroom costs from the 6% definition does not protect classroom services from all effects of a revenue limit; effects depend on each budget and any override.

[55] [56]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The law does exclude classroom instruction and school-based services from the central-office definition. That does not remove the indirect effect of a limit on district tax revenue. Nor does it establish that every district will cut classrooms: costs, other aid, reserves, efficiencies and overrides determine the outcome.[55] [56]

Check the calculation

Illustration, not a forecast: non-bond local levy $10m; inflation 3%; net new taxable growth 1% → next-year limit $10.4m. If preserving services requires $10.8m after all other revenues, the $0.4m difference must be resolved through an override, lower spending or other available resources. A 6% administrative limit alone does not resolve it.

[55] [56]

Source trail

02BallotDifferent numerator and denominator

Census “administration” is not the 6% legal category.

Census combines general and school administration in the $2,140 figure. The ballot law excludes school-based administrative staff and uses an appropriation-based denominator with bonded-capital exclusions.[1] [55]

Ballot context

$2,140 per pupil ≠ a 6% SAU-cap calculation

Context: the verified administration graphic is relevant to cost scrutiny, but it cannot directly measure whether an SAU exceeds the ballot cap.

Read the enacted law ↗
What the primary record shows

RSA 32:5-i IX covers superintendent services, business administration, HR, finance, legal and certain district-level instructional leadership functions, regardless of physical office location. It expressly excludes school-based administrative staff, classroom instruction and listed services.[1] [55]

What this means, in precise terms

The $2,140 administration figure is accurate, but it includes principal offices and uses a different spending base from Question 2. It cannot establish that an SAU is over the 6% limit. That requires the statute’s specific costs and appropriation denominator.

[1] [55]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

Census’s general-administration category is not automatically identical to the statute either. Some district-level services may be classified differently. A payroll/function crosswalk and appropriation schedule are required. Comparing 9.31% of current spending with a 6% cap on a differently defined appropriation base is not valid.[1] [55]

Check the calculation

FY2024 Census: general $848.655 + school $1,291.073 = $2,139.728 per pupil. Combined ÷ current spending $22,978.148 = 9.31%; general alone = 3.69%. Neither percentage is the statutory cap test. The law’s denominator is the sum of member-district appropriations excluding certified bonded capital.

[1] [55]

Source trail

03BallotSeparate legal questions

A vote on the cap does not overturn the funding ruling.

ConVal concerns the state’s funding obligation. HB1300 creates local votes on tax and central-office limits. The September 2026 Supreme Court order denied a petition for original jurisdiction about the ballot.[11] [55]

Ballot context

ConVal 2025 • HB1300 2026 • Bates petition 2026

Context: three different legal events are sometimes blended into one story.

Read the enacted law ↗
What the primary record shows

The July 1, 2025 ConVal opinion affirmed inadequacy and a conservative minimum guidepost, but reversed immediate-payment relief. HB1300 was signed July 15, 2026. On September 28, the Court’s one-page Bates order denied original jurisdiction under Rule 11 without substantive explanation.[11] [55] [56] [57]

What this means, in precise terms

The funding ruling did not declare school expense increases unconstitutional. It found the challenged state base-funding law inadequate. Question 2 is a separate local cap vote, and the brief 2026 order allowing the ballot process to continue is not a finding that the caps cannot affect services.

[11] [55] [56] [57]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The 2026 order does not say that classroom services are guaranteed unaffected. The local ballot does not itself amend the constitution, reverse ConVal, appropriate a larger adequacy grant or decide all later funding litigation. The chronology shows related debates, not proof of a single motive or coordinated campaign.[11] [55] [56] [57]

Check the calculation

If a district’s local revenue is constrained while the state duty remains unmet, the financing tension remains. That is a conditional policy consequence, not a prediction of the result in every district. Keep the state-duty question, tax-limit formula and legal procedural outcome separate.

[11] [55] [56] [57]

Source trail

04StaffingSupported with context

The administration number is real. Its scope matters.

The Census comparison checks out. It includes principal offices as well as district-level administration; treating all of it as removable SAU overhead overstates the opportunity.[1] [4]

Claim being checked

$2,140; 65%; $849; 2.4

FY2024 administration costs about $2,140 per pupil, 65% above the U.S.; general administration alone is about $849 versus $351.

View related original graphic ↗
What the primary record shows

FY2024 Table 8 reports NH general administration $848.655 and school administration $1,291.073 per pupil. U.S. counterparts are $351.270 and $945.716.[1] [4]

What this means, in precise terms

The Census administration comparison is accurate: about $2,140 per pupil in FY2024, 65% above the U.S. It includes principal offices, so it is not a measure of SAU overhead alone.

[1] [4]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

These categories include salaries, benefits and other costs—not just superintendent pay. Eliminating an office does not eliminate all its legally required work. The interstate gap does not, by itself, identify waste.[1] [4]

Check the calculation

NH: 848.655 + 1,291.073 = $2,139.728. U.S.: 351.270 + 945.716 = $1,296.986. Premium: (2,139.728 ÷ 1,296.986 − 1) × 100 = 64.98%. General ratio: 848.655 ÷ 351.270 = 2.416. Combined administration is 9.31% of NH current spending ($22,978.148), not most of the budget.

[1] [4]

Source trail

05StaffingCounts verified; savings unproven

A 1993 estimate is not a current savings forecast.

The SAU counts are independently reproduced. The historical savings estimate is real as a documented estimate, but it is not a current demonstrated saving.[48] [3]

Claim being checked

101; 50; $62 million

The site counts 101 SAUs, 50 below 1,000 students, and cites $62 million in possible annual savings from replacing SAUs.

View related original graphic ↗
What the primary record shows

The final NHED report contains 203 unique district rows totaling 160,322 pupils. Grouping rows with SAU numbers below 200 yields 101 SAUs; 50 have fewer than 1,000 pupils. The 2025 committee report separately cites the historical 1993 savings model.[48] [3] [4]

What this means, in precise terms

NHED’s October 2025 report does show 101 SAUs, with 50 below 1,000 students. That makes scale and shared services legitimate questions. But the $62m saving is a restated 1993 estimate: presenting it as money a current reorganization would deliver is unsupported.

[48] [3] [4]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The current SAU counts are verified. The $62m is a purchasing-power restatement of an estimated $27m saving in a 1993 service model—not a costed estimate of current county restructuring. Required duties, transition spending and potential salary harmonization still need analysis.[48] [3] [4]

Check the calculation

Historical restatement: $62m ÷ $27m ≈ 2.30 inflation multiplier. It changes the purchasing-power unit; it does not update the staffing/service model or demonstrate a $62m cash saving in 2026.

[48] [3] [4]

Source trail

06StaffingNumbers verified; meaning narrower

More staff per pupil does not mean more administrators.

The NCES staff ratio is reproducible. It covers all staff, including teachers and aides. It is not an administrator count or a class-size measure.[2] [49]

Claim being checked

18.5; 13.7; 35%; +10%

NH has about 35% more public-school staff per student than the U.S.; staffing per student rose about 10% over 2012–2022.

View related original graphic ↗
What the primary record shows

NCES reports NH 31,198 full-time-equivalent staff and 168,909 pupils in fall 2022, compared with 31,734 FTE and 188,974 pupils in fall 2012. U.S. fall 2022 totals are 6,795,470 FTE and 49,618,464 pupils.[2] [49]

What this means, in precise terms

The 35% NH/U.S. staffing gap and 10% decade increase per pupil check out. But this is all staff, not administrators. NH’s FTE staff count fell about 1.7% from 2012 to 2022 while enrollment fell about 10.6%; the ratio rose because pupils declined faster.

[2] [49]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The counts cover all staff, including teachers and aides, not just administrators. FTE staff actually fell 1.69% across the decade; staff per pupil rose because enrollment fell faster. A ratio is not class size or evidence that each job is unnecessary.[2] [49]

Check the calculation

NH 2022: 31,198 ÷ 168,909 × 100 = 18.470 FTE per 100 pupils. U.S.: 13.695. NH premium = 34.86%. NH 2012: 16.793 FTE per 100. Relative growth = 9.99%; NCES enrollment change = −10.62%.

[2] [49]

Source trail

07StaffingStaffing pattern independently verified

Job categories are not interchangeable.

The major staffing-growth percentages are reproduced from NCES. These job categories do not all describe administration.[50] [51]

Claim being checked

−14%; +2%; +40%; +41%; +57%; +61%

The site contrasts falling enrollment with increases in teachers, aides, student support, coordinators and district administrators over school years ending 2001–2019.

View related original graphic ↗
What the primary record shows

The corresponding NCES school-year endpoints are fall 2000 and fall 2018. Teachers: 14,341→14,644 (+2.11%); aides 5,056→7,061 (+39.66%); district officials/administrators 439→690 (+57.18%); coordinators 173→278 (+60.69%); student support 542→767 (+41.51%).[50] [51] [4]

What this means, in precise terms

The staffing pattern is real, and staffing choices deserve scrutiny. But added aides and student-support staff cannot fairly be described as added SAU bureaucracy. Percent changes also need their starting counts: the 57% district-administrator increase means 251 FTE in these tables.

[50] [51] [4]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

Aides and student support are separate from district administration. A large percentage increase from a small base does not establish the budget share or whether the jobs are needed. The small 41% versus 41.51% difference reflects the inspected publication/rounding; the separate pay comparison is checked in its own card below.[50] [51] [4]

Check the calculation

For each category: (fall 2018 FTE ÷ fall 2000 FTE − 1) × 100. District administrators rise by 251 FTE, from 439 to 690; that is 57.18%. These endpoints correspond to school years ending 2001 and 2019.

[50] [51] [4]

Source trail

08StaffingArithmetic reproduced; spending definition changes

Teacher pay grew much less than total spending.

The 8% / 45% figures reproduce mathematically. Pay growth is 7.61% under the original inflation convention. But the 44.57% spending calculation adds debt principal at the end of the series while excluding it at the start.[68] [77]

Claim being checked

Real teacher pay +8%; real total spending +45%, 2001–2024

The site contrasts growth in average teacher salary with total district spending and interprets the difference as a choice to prioritize staffing.

View related original graphic ↗
What the primary record shows

The original expenditure baseline also reproduces from NCES: FY2001 TE11 $1,690,009,689 plus E7A1 interest $33,014,855 equals $1,723,024,544. The original salary endpoints are $38,301 in 2000–01 and $67,170 in 2023–24. January PCE is 74.932 in 2001 and 122.115 in 2024 in the April 2025 vintage. That exact ratio turns the report’s $1,723,024,544 starting expenditure into $2,807,974,459 in 2024 dollars—the printed chart baseline to the nearest dollar. FY2024 spending is $3,980,451,955 plus $79,041,383 debt principal, totaling $4,059,493,338.[68] [77] [76] [75] [34] [35] [69] [62] [65] [67] [61] [82]

What this means, in precise terms

The 8% / 45% figures reproduce, but the spending series changes definition. Its 2001 baseline excludes debt principal, while the 2024 endpoint adds it. Removing that addition changes 45% to about 42% under the same inflation method. Teacher pay still grew much less than total spending. That contrast deserves scrutiny, but it does not itself identify waste or explain every additional dollar.

[68] [77] [76] [75] [34] [35] [69] [62] [65] [67] [61] [82]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The spending comparison changes definition: FY2001 uses NCES total expenditure including debt interest but excluding principal; FY2024 adds $79.04m of principal to NHED expenditure. Removing just that endpoint addition lowers the same calculation from 44.57% to 41.76%, before addressing other coverage differences. The original inflation factor also uses January observations, not school-year averages. Average gross salary and total spending measure different things: benefits, other staff, services and capital are in the spending total. FY2024 omits a nonreporting district; salary and expenditure populations differ. The broad pay/spending contrast survives, but 45% is not a clean growth rate on one unchanged expenditure definition, and neither comparison alone identifies unnecessary hiring.[68] [77] [76] [75] [34] [35] [69] [62] [65] [67] [61] [82]

Check the calculation

Original PCE factor = 122.115 ÷ 74.932 = 1.629677574. Salary: ($67,170 ÷ $38,301) ÷ factor − 1 = 7.612704%, rounding to 8%. Restated starting expenditure: $1,723,024,544 × factor = $2,807,974,459.38. Spending: ($4,059,493,338 ÷ $1,723,024,544) ÷ factor − 1 = 44.570166%, rounding to 45%. The current revised January 2024 PCE value changes spending to 44.46%; this is a data-vintage effect. Separately, July–June current PCE gives +7.27% salary. Fiscal-year CPI-U gives −0.81% salary; neither consumer index measures school input costs. Definition sensitivity: replace $4,059,493,338 with NHED expenditure before principal, $3,980,451,955. With the same baseline and PCE factor, growth is 41.755276%. This isolates the principal addition; it does not eliminate all other population or accounting differences.

[68] [77] [76] [75] [34] [35] [69] [62] [65] [67] [61] [82]
The original calculation • exact inputs and rounding
MeasureStartEnd / result
Mean teacher salary$38,301$67,170
PCE, January observations; April 2025 vintage74.932122.115
Report’s total cash-outlay measure$1,723,024,544$4,059,493,338
2001 spending restated into 2024 dollars$1,723,024,544 × 1.629677574$2,807,974,459.38; exact chart match
Real salary growth7.612704%Rounds to 8%
Real total spending growth44.570166%Rounds to 45%

Inflation sensitivity, using the same salaries: current July–June PCE gives +7.27% real salary growth; July–June CPI-U gives −0.81%. These are explicitly different conventions, not replacement numbers used to mark the original calculation wrong.

A material definition change: NCES’s starting total excludes debt principal. The original report adds $79,041,383 of principal to the FY2024 endpoint. Without that addition, growth is 41.76% under the same inflation calculation. Reproducing 45% verifies the arithmetic, not the comparability of the expenditure definitions.

Source trail

09SpendingHeadline arithmetic verified

The spending trend deserves scrutiny, not a shortcut.

The stated rounded spending increase and enrollment decline are now verified against the recovered original records. The interpretation still depends on aligned years and pupil coverage.[6] [8]

Claim being checked

+82%; −21%; 48,139

The site reports inflation-adjusted per-pupil cost up 82% from 2000 to 2023, enrollment down about 21%, and a 48,139 decline through fall 2025.

View related original graphic ↗
What the primary record shows

The recovered NHED chart shows $11,192 in 1999–2000, converted to 2023 dollars, and $20,323 in 2022–23. NCES confirms 208,461 pupils in fall 2000; live NHED reports sum to 165,082 in fall 2023 and 160,322 in fall 2025.[6] [8] [46] [47]

What this means, in precise terms

The site’s rounded 82% real cost-per-pupil increase and 48,139 long-run enrollment decline are supported by the stated source endpoints. The spending and fall-enrollment years do not exactly align. The trend justifies scrutiny; it does not establish which services can be removed without consequences.

[6] [8] [46] [47]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The years are slightly misaligned: FY2000 cost precedes fall 2000, and FY2023 cost precedes fall 2023. Fall counts, ADM-A, charter coverage and grade coverage differ across sources. The verified rise does not prove that every current program is sufficiently funded or identify a safe cut.[6] [8] [46] [47]

Check the calculation

Reported cost change: (20,323 ÷ 11,192 − 1) × 100 = 81.59%. Enrollment: (165,082 ÷ 208,461 − 1) × 100 = −20.81%; 208,461 − 160,322 = 48,139. Exact total real-spending change needs expenditures and membership aligned by fiscal year; do not multiply mismatched series.

[6] [8] [46] [47]

Source trail

10SpendingNumbers verified; definitions matter

Use the same cost definition on both sides.

Both the FY2024 Census comparison and FY2025 NHED figures are now independently verified. They measure different costs and periods, so the labels are essential.[5] [41]

Claim being checked

$22,700; $27,690; $22,978; $17,619; $5,359

The site reports NHED FY2025 operating and broader cost measures, and a $5,359 FY2024 current-spending gap over the U.S.

View related original graphic ↗
What the primary record shows

NHED FY2025 reports $22,699.85 per ADM-A in operating expenses, $27,689.90 for total elementary/secondary expenditures, and $27,829.65 when all listed non-K–12 programs are included. Census FY2024 independently confirms NH $22,978.148 current spending versus U.S. $17,619.389.[5] [41] [1]

What this means, in precise terms

NH’s reported costs are high, and the $5,359 FY2024 Census gap is correct. But “cost per pupil” can mean $22,699.85 in FY2025 operating expenses, $27,689.90 for elementary/secondary totals, or $27,829.65 including all listed programs. A comparison can change materially when its cost definition changes.

[5] [41] [1]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The $27,690 figure is a valid elementary/secondary measure, but it is not the agency table’s grand all-program figure. The district PDF’s introductory wording and the detailed state table are not perfectly consistent on that point; use the detailed table. NHED and Census use different coverage/denominators and different fiscal years here.[5] [41] [1]

Check the calculation

FY2025: $3,391,032,564 operating ÷ 149,385.66 ADM-A ≈ $22,699.85. Elementary/secondary $4,136,474,047 ÷ ADM-A ≈ $27,689.90. Grand total $4,157,349,112 ÷ ADM-A ≈ $27,829.65. FY2024 Census gap = $5,358.759, or 30.41%.

[5] [41] [1]

Source trail

11SpendingCounts independently reproduced

The district counts check out. Their years differ.

The original NHED chart and fall-enrollment records support 158 increases and 139 enrollment declines in that comparison group. The two tests still use different periods.[6] [44]

Claim being checked

158 of 158; 139 of 158

The site says all 158 comparable districts spent more per pupil after inflation; 139 lost students.

View related original graphic ↗
What the primary record shows

The four-page cost chart contains 158 district rows with two amounts, all increasing after inflation. Four repeated state-average rows were excluded; two rows missed by the first OCR grouping were visually verified. Joining the same 158 districts to NHED fall 2011 and fall 2025 reports yields 139 declines.[6] [44] [47]

What this means, in precise terms

The 158-of-158 spending and 139-of-158 enrollment counts can be reproduced from NHED. The key qualification is time: cost is compared from FY2000 to FY2023, while enrollment runs from fall 2011 to fall 2025. Both patterns are real; combining them does not establish what a current cut would do.

[6] [44] [47]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

These are the comparable districts, not every reporting entity. The cost comparison is 1999–2000 to 2022–23, while the enrollment comparison is fall 2011 to fall 2025. Legacy reports reuse some identifiers for other entities, so district names as well as IDs were used to disambiguate rows. The counts describe a trend, not the effect of a proposed cut.[6] [44] [47]

Check the calculation

Cost: 158 of 158 directly paired district rows increase. Enrollment: 139 ÷ 158 = 87.97%, or nearly nine in ten. Pair the exact same district set across enrollment years; do not sum duplicate print representations or treat a reused ID as sufficient identity.

[6] [44] [47]

Source trail

12FundingConclusion not established

A spending concern and a funding failure can coexist.

Efficiency and constitutional adequacy are separate questions. High statewide spending does not establish that the state has funded the educational opportunity it is required to provide.[1] [11]

Claim being checked

“a spending problem, not a funding problem”

The site uses spending growth to reject a funding problem.

View related original graphic ↗
What the primary record shows

NH spends more per pupil than the U.S. average, and administrative costs warrant review. It is fair to ask whether dollars deliver effective services.[1] [11] [9]

What this means, in precise terms

NH can have spending inefficiencies and a state-funding problem at the same time. The 2025 ConVal ruling found the challenged base-aid law inadequate; a statewide spending average does not resolve that legal question.

[1] [11] [9]
Calculation, qualifications & source trail A + interpretation

Where the framing can mislead—or the evidence stops

ConVal’s July 1, 2025 ruling affirmed that the challenged base-aid law was facially unconstitutional and upheld a conservative minimum guidepost exceeding $7,356.01 under the existing framework. The immediate-payment directive was reversed. This is a dated holding, not verification of all subsequent proceedings or remedies through October 2026.[1] [11] [9]

Check the calculation

A statewide expenditure average measures actual spending from multiple sources. A constitutional adequacy inquiry asks what opportunity the state must fund and at what cost. Neither number substitutes for the other.

[1] [11] [9]

Source trail

13FundingFY2025/FY2026 rates verified

Base aid is neither the full state contribution nor actual cost.

The base is one component of the formula. Calling it all state education funding understates funding; calling the full spending average proof of state adequacy avoids the legal issue.[9] [43]

Claim being checked

$4,182; $4,266

The site cites base adequacy aid of $4,182 in FY2025 and $4,266 in FY2026.

View related original graphic ↗
What the primary record shows

NHED’s original FY2026 explanation gives $4,265.64 per resident ADM, rounded by the site to $4,266. The recovered EFA fact sheet’s historical rate table confirms FY2025 $4,182. These are the base component, before differentiated aid and other funding.[9] [43] [22] [1] [13]

What this means, in precise terms

The cited base rates—$4,182 in FY2025 and $4,265.64 in FY2026—are verified. They are not all state support, and total spending is not the same as the cost of the state’s constitutional obligation. Treating either as the other misframes the funding question.

[9] [43] [22] [1] [13]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The FY2026 explanation was issued in November 2024 at estimate stage; subsequent amendments and final awards need their own dates. A 2025–26 rate is not a current FY2027 rate in October 2026. The base, broad state-source revenue, SWEPT credit and actual cost are different quantities.[9] [43] [22] [1] [13]

Check the calculation

Verified FY2024 Census state-source revenue: $1,144,508,000 ÷ 156,325 pupils = $7,321.34 per pupil. That broad state-source revenue measure is not base aid and includes state-source classifications such as SWEPT; it is also not a FY2025 aid rate.

[9] [43] [22] [1] [13]

Source trail

14Taxes62% share and 103% growth verified

A tax share is not an explanation of a tax increase.

The rounded 62% school share and 103% nominal local-school tax increase are independently reproduced. Neither figure explains what caused the increase or predicts an individual bill.[30] [42]

Claim being checked

62%; $3.15B; $5.09B; $1.38B → $2.79B

The site says schools receive 62% of property taxes and local school taxes rose 103% from 2007–08 to 2025–26.

View related original graphic ↗
What the primary record shows

All 260 rows of the completed DRA report yield total commitments of $5,092,750,972. Rate-times-value reconstruction gives local school $2,789,647,675 and SWEPT $363,383,292, together $3,153,030,967. This is 61.912% of commitments, supporting the rounded 62%. NHED’s October 2026 revision separately totals local school assessment at $2,790,227,910.[30] [42] [60] [52] [61] [31] [13]

What this means, in precise terms

The school share rounds to 62%, and original NHED tables confirm local-school assessments rose about 103% in nominal dollars from 2007–08 to 2025. After the verified 55.09% national consumer-price rise, aggregate real growth is about 30.62%. Calling that a 103% increase in the real tax burden would be misleading: the number of taxpayers, the tax base and individual assessments also change. The totals alone do not identify the cause.

[30] [42] [60] [52] [61] [31] [13]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

Rounded-rate component estimates are gross and do not exactly reconcile to net commitments; a same-gross-basis share is 61.752%, also rounding to 62%. The original NHED 2007–08 and 2025 tables independently confirm the 103% nominal local-school increase. That is not an inflation-adjusted increase, a per-household increase or a measure of the effect of one policy.[30] [42] [60] [52] [61] [31] [13]

Check the calculation

Original NHED totals: ($2,790,227,910 ÷ $1,377,385,153 − 1) × 100 = 102.574%, rounding to 103%. Separately, DRA rate × applicable value ÷ 1,000 must be summed across all report rows; utilities are excluded for SWEPT but included for local school. BLS national CPI-U (not seasonally adjusted) rises from 208.299 in July 2007 to 323.048 in July 2025: +55.0886%, confirming the rounded 55%. Real local-school growth = (2.02574269 ÷ (323.048 ÷ 208.299) − 1) × 100 = 30.62%. This is a general consumer-price adjustment, not a school-specific cost index.

[30] [42] [60] [52] [61] [31] [13]

Source trail

15FundingOriginal figures reproduced; coverage mismatch

Local funding grew much faster; the original denominators do not match.

The −3% state / +135% local figures now reproduce exactly using a 170-district revenue panel and broader statewide pupil counts. That is a reconstructed method; the report does not explicitly disclose the 170-district join. Using membership from the same panel changes the result to about −0.27% state / +142.22% local.[83] [81]

Claim being checked

State −3%; local +135% per pupil, 2001–2019

The site quotes an older policy report’s inflation-adjusted state funding of $5,791→$5,604 per pupil and local funding of $5,223→$12,279.

View related original graphic ↗
What the primary record shows

Match the FY2001 and FY2019 federal district-finance records by district identifier and retain revenue data present in both years: 170 districts. Their state revenue is $877,999,000→$1,000,427,000; local revenue $792,004,000→$2,191,912,000. With the original January PCE factor and statewide pupils 208,461→178,515, the real per-pupil amounts round to state $5,791→$5,604 and local $5,223→$12,279. Every displayed endpoint and rounded percentage is recovered.[83] [81] [85] [34] [82] [62] [63] [13] [84]

What this means, in precise terms

The older state/local funding numbers are traceable, but their populations do not line up. Revenues come from districts present in both years; pupils come from broader statewide totals. Matching the pupil count to that revenue panel gives roughly flat real state funding per pupil and a 142% rise in local revenue. The central contrast survives. The precise 3% state decline depends on the original denominator choice.

[83] [81] [85] [34] [82] [62] [63] [13] [84]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The revenue panel and pupil denominator have different coverage. The same 170 records contain 158 districts with positive membership in both years, totaling 203,448→169,074 pupils, plus 12 without positive membership. Using that panel’s counts makes real state revenue per pupil approximately flat, while local revenue still rises sharply. This is a coverage-aligned sensitivity, not a claim that every financing measure must use one universal denominator. The original local category is total local-source revenue, broader than property taxes. These trends do not identify how much tax revenue would have been needed without state aid.[83] [81] [85] [34] [82] [62] [63] [13] [84]

Check the calculation

Historical PCE factor = January 2019 108.977 ÷ January 2001 79.265 = 1.3748438781303223, using the March 2023 data vintage. Original state: ($1,000,427,000 ÷ 178,515) ÷ (($877,999,000 ÷ 208,461) × factor) − 1 = −3.21945%. Original local: apply the same formula to $792,004,000→$2,191,912,000 = +135.06743%. Change only the pupil counts to 203,448→169,074: state −0.27258%; local +142.22499%. Converting the raw sums to 32-bit floating-point values reproduces the printed aggregate totals. The responsible software is not established; the less-than-$65 precision differences do not affect the rounded findings.

[83] [81] [85] [34] [82] [62] [63] [13] [84]
Original source reconstruction • same 170 district-finance records
InputFY2001FY2019
Primary state-revenue sum$877,999,000$1,000,427,000
Primary local-revenue sum$792,004,000$2,191,912,000
Original broad statewide pupil count208,461178,515
Pupils in the matched finance panel203,448169,074
Original January PCE; March 2023 vintage79.265108.977
What changes when the population is aligned • 2019 dollars per pupil
CalculationState revenue / pupilLocal revenue / pupil
Original broad pupil denominator$5,790.59 → $5,604.16 (−3.22%)$5,223.43 → $12,278.59 (+135.07%)
Same-panel pupil denominator$5,933.27 → $5,917.10 (−0.27%)$5,352.14 → $12,964.22 (+142.22%)

The two district samples are not competing totals: 170 districts supply the matched revenue panel; 158 have positive membership in both years. The publisher’s district appendix uses those 158. The report’s statewide total-expenditure series comes from a separate NCES state-finance definition. Matching one table’s total to a different table was the reason the earlier audit missed the original method.

Reconstruction, not an unpublished author instruction: the 170-district join is supported by an exact match across state, local and federal revenues, together with the publisher’s appendix membership and inflation factor. The report does not explicitly state that selection rule. The precise software stage responsible for the immaterial floating-point differences is also unknown.

Source trail

16FundingAll six original figures reproduced

Recent funding growth changes after inflation.

The original NHED release reproduces all six per-pupil amounts and the rounded increases of 19%, 29% and 126%. The earlier discrepancy came from using a later financial release and omitting “other” local revenue.[53] [75]

Claim being checked

State +19%; local +29%; federal +126%, 2020–2024

The site’s underlying report quotes these nominal per-pupil increases, alongside roughly 17% inflation, and questions whether additional state aid reduces taxes.

View related original graphic ↗
What the primary record shows

FY2020: state $6,295.87, local $12,761.49 and federal $965.16 per pupil. FY2024: $7,468.29, $16,399.88 and $2,179.34. These round to every value in the report. Its fall-enrollment denominators are 167,584 and 154,080; FY2024 excludes a district without financial reporting. The table below shows the exact revenue components. Both denominators are now independently verified: the 2019 fall report totals 167,584 traditional-district pupils; the 2023 report totals 155,253, minus both rows carrying the excluded financial district’s identifier (1,019 + 154), equals 154,080.[53] [75] [35] [76] [64] [65] [66] [86] [46]

What this means, in precise terms

The recent financing figures check out exactly against the original NHED release. State, local and federal revenue per pupil rose about 19%, 29% and 126% in nominal dollars. Inflation accounts for much of the state increase. These trends do not establish that state aid fails to reduce the tax burden: the missing comparison is what taxes would have been without the aid.

[53] [75] [35] [76] [64] [65] [66] [86] [46]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The local category includes tuition/food receipts and other revenue; it is not all property tax. The report changes coverage in FY2024, so this is not a fixed-district panel. January-to-January PCE inflation of 16.95% reduces the nominal increases to about 1.43%, 9.89% and 93.08% in real per-pupil terms. Rising aid and rising taxes together do not show whether taxes would have been higher without the aid.[53] [75] [35] [76] [64] [65] [66] [86] [46]

Check the calculation

State FY2020: ($961,613,351 + $93,473,735) ÷ 167,584 = $6,295.870. State FY2024: ($1,056,696,504 + $94,018,279) ÷ 154,080 = $7,468.294. Local FY2020: ($2,099,195,588 + $39,279,701 + $145,685) ÷ 167,584 = $12,761.487. Local FY2024: ($2,446,703,088 + $67,622,977 + $12,567,723) ÷ 154,080 = $16,399.882. Federal: $161,745,104 ÷ 167,584 = $965.158; $335,792,202 ÷ 154,080 = $2,179.337. Change = ending per-pupil amount ÷ starting amount − 1; divide that ratio by 122.115 ÷ 104.419 for the real change.

[53] [75] [35] [76] [64] [65] [66] [86] [46]
Original report reconstructed • nominal dollars, before inflation
Revenue / denominatorFY2020FY2024
State sources$1,055,087,086$1,150,714,783
Local property tax$2,099,195,588$2,446,703,088
Local tuition / food$39,279,701$67,622,977
Other local receipts$145,685$12,567,723
Total local used in report$2,138,620,974$2,526,893,788
Federal sources$161,745,104$335,792,202
Report’s fall-enrollment denominator167,584154,080
Rounded state per pupil$6,296$7,468
Rounded local per pupil$12,761$16,400
Rounded federal per pupil$965$2,179

Why the earlier check differed: the January 7, 2025 NHED release is the appropriate source for the older report. A revised January 2026 release and NCES’s finance-adjusted pupil counts answer related but different questions. The small “other” revenue category must also be included to reproduce local funding.

The last denominator discrepancy: the 2023 enrollment report lists two rows with the excluded financial district’s identifier. Removing only its main named row gives the wrong result. Removing both gives 155,253 − 1,019 − 154 = 154,080, exactly the original report’s denominator. The FY2020 comparison includes that district; the FY2024 comparison excludes it.

Source trail

17HistoryVerified policy history

Retirement cost shifting is documented policy.

State policy changed who pays the employer pension bill. That is a concrete cost shift, although it does not explain all school-spending growth.[19] [16]

Claim being checked

Context omitted from the share graphics

The share graphics do not explain the historical state retirement contribution reductions.

View related original graphic ↗
What the primary record shows

The LBA records a former 35% state share of eligible local employer contributions, reduced to 30% in FY2010 and 25% in FY2011. The 2011 law substituted $3.5m in FY2012 and eliminated the payment in FY2013. A 2022 law provided one-time aid, with $27.1m spent in FY2023.[19] [16] [17] [18] [20]

What this means, in precise terms

The state reduced and then ended its historical share of eligible local employer retirement contributions, with elimination in FY2013. That shifted responsibility locally; it was not a 35% subsidy of the entire school budget.

[19] [16] [17] [18] [20]
Calculation, qualifications & source trail A

Where the framing can mislead—or the evidence stops

This is 35% of eligible employer contributions, not 35% of salaries or school budgets. The LBA also records the legislature’s rationale that retirement reforms would offset costs; that rationale is not independent proof of a full offset. Current contribution rates and later aid require current records.[19] [16] [17] [18] [20]

Check the calculation

Illustration, fixed eligible employer contribution bill of $10m: a 35% state share is $3.5m; local share $6.5m. Removing the subsidy makes local share $10m, an increase of $3.5m or 53.85% of its former share. This is an illustrative constant-bill calculation, not measured statewide impact.

[19] [16] [17] [18] [20]

Source trail

18TaxesCorrect classification; important omitted context

SWEPT is state tax policy collected on property.

SWEPT belongs in property-tax totals, and it is a state-imposed tax. Counting it as state-source school revenue does not make it cease being a property-owner payment.[13] [15]

Claim being checked

$363,000,000

The site counts statewide education property tax in the school share of the property-tax bill.

View related original graphic ↗
What the primary record shows

RSA 76:3 sets a statewide revenue target of $363m. The 2005 law established the target. State-source and local-source labels answer an accounting question; the taxpayer’s bill answers who pays.[13] [15] [1]

What this means, in precise terms

SWEPT is a state-imposed property tax with a statutory $363m target. It may be classified as state revenue while still being paid on local property-tax bills; count it once.

[13] [15] [1]
Calculation, qualifications & source trail A

Where the framing can mislead—or the evidence stops

A fixed nominal target does not grow automatically with school costs or property values. This does not by itself prove total state aid fell: other state revenues and grants can increase. Do not count SWEPT twice when totaling resources.[13] [15] [1]

Check the calculation

For illustration only, if prices double while a nominal $363m target stays fixed, its purchasing power becomes $181.5m in starting-year dollars. That is a scenario, not the measured 2005–2026 inflation change.

[13] [15] [1]

Source trail

19History2025 holding verified

Excess retention: distinguish legality from fairness.

The June 10, 2025 Rand majority upheld excess retention under Part II, Article 5. It separately rejected negative local education tax rates. Calling excess retention categorically unconstitutional misstates that decision.[12] [13]

Claim being checked

Context absent from the statewide share cards

The site’s tax graphics do not resolve the history or distribution of excess SWEPT.

View related original graphic ↗
What the primary record shows

The opinion records retention since 2011 and holds that uniform collection with local retention is a legislative spending decision, not the asserted tax-uniformity violation. The dissent disputes that reasoning.[12] [13] [9]

What this means, in precise terms

In June 2025, the NH Supreme Court upheld excess SWEPT retention under the tax-uniformity provision and rejected negative local education tax rates. Fairness and adequacy remain separate questions.

[12] [13] [9]
Calculation, qualifications & source trail A

Where the framing can mislead—or the evidence stops

A constitutional ruling on that issue is not a finding that the distribution is fair, adequate or economically neutral. The opinion does not resolve every school-funding claim or certify later implementation. More taxable wealth per pupil can produce excess even at a uniform rate.[12] [13] [9]

Check the calculation

At a common state rate r, SWEPT per pupil = r × equalized taxable wealth ÷ pupils. A uniform rate therefore does not generate uniform revenue per pupil. Cash adequacy aid accounts for credited SWEPT under the formula.

[12] [13] [9]

Source trail

20EFAsMixed years; later same-year projection available

The penny graphic compares different fiscal years.

The recovered inputs confirm the graphic’s roughly 1.24% calculation. It combines a later projection with earlier actual spending, and the source supplies a newer same-year projection than the initial $27.7m figure.[22] [41]

Claim being checked

$51.6 million; $4.16 billion

The share graphic compares projected FY2026 EFAs with FY2025 district spending, presenting the result as about one cent per education dollar.

View related original graphic ↗
What the primary record shows

The original November 2025 NHED fact sheet projects FY2026 EFA awards at $51,619,550. Its retrospective FY2025 beginning-year projection is $30,354,282. NHED’s FY2025 district expenditure total is $4,157,349,112. These inputs are independently recovered.[22] [41] [21]

What this means, in precise terms

“About a penny” is a rough scale comparison, not an audited same-year budget share. The graphic mixes FY2026 projected EFA awards with FY2025 district spending. A later NHED projection for FY2025 is $30.35m—about 0.73% of that year’s $4.157b district expenditure—not the earlier $27.7m estimate.

[22] [41] [21]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The $51.6m graphic uses a FY2026 projection over FY2025 actual district spending. The earlier $27.7m initial estimate is not the only available FY2025 estimate: the later fact sheet gives $30.35m. Neither award projection is an audited measure of actual family spending or a measure of local marginal impact.[22] [41] [21]

Check the calculation

Mixed-year ratio: $51,619,550 ÷ $4,157,349,112 = 1.2416%. Aligned fiscal-year comparison, still projected awards versus actual spending: $30,354,282 ÷ $4,157,349,112 = 0.7301%. Including both in the denominator gives 0.7248%. None is a final actual-award/outlay ratio.

[22] [41] [21]

Source trail

21EFAsAmounts verified; savings inference misframed

A smaller grant is not automatically a taxpayer saving.

A smaller EFA award than average public-school expenditure is a verified observation. It is not, on its own, a verified saving to taxpayers.[21] [22]

Claim being checked

$4,911; $22,700

The site contrasts a projected FY2026 EFA average with FY2025 public-school operating cost, implying EFAs cost taxpayers less.

View related original graphic ↗
What the primary record shows

The November 2025 fact sheet verifies the projected FY2026 average EFA award of $4,911 (total projected awards $51,619,550 for 10,510 recipients). The FY2025 NHED operating average is $22,699.85. The grant is genuinely smaller than that average, but the comparison uses different years and different cost concepts.[21] [22] [5] [41]

What this means, in precise terms

The $4,911 EFA award and $22,699.85 public operating average are verified, but subtracting them does not establish taxpayer savings. The comparison mixes fiscal years and substitutes average cost for avoidable cost. The actual effect depends on the student’s alternative and which public costs disappear.

[21] [22] [5] [41]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

An average grant is not the same as the cost a public school can avoid when one pupil leaves. A pupil already privately educated may create a new public subsidy; a departing public pupil may leave fixed costs behind. The two populations also have different service responsibilities and eligibility profiles. Longer-run adjustments may produce savings.[21] [22] [5] [41]

Check the calculation

Illustration only: if avoidable public costs are $2,000 and a grant is $5,000, combined public cost rises $3,000 before other effects. If avoidable costs are $8,000, it falls $3,000. District levy pressure depends separately on aid lost minus costs avoided minus transition support. Do not add state-aid loss to the grant again when calculating consolidated taxpayer cost.

[21] [22] [5] [41]

Source trail

22EFAsCounts verified; causal framing limited

EFAs did not cause the pre-2021 enrollment decline.

The timing and reported transfer arithmetic check out. The date range and denominators do not turn those ratios into causal estimates of the program’s effect.[23] [22]

Claim being checked

84%; 1,437; 3%

The site says 84% of the long-term enrollment decline preceded EFAs and reported district-to-EFA transfers equal about 3% of the decline since 2000.

View related original graphic ↗
What the primary record shows

NCES/NHED endpoints confirm 208,461 in fall 2000, 167,909 in fall 2020 and 160,322 in fall 2025. EFA Table 12 reports 275, 213, 309, 297 and 343 prior-district-school entrants for school years ending 2022–2026. The last two figures are preliminary; the table describes phase-out-grant counts, not all conceivable program-induced enrollment effects.[23] [22] [45] [47] [8]

What this means, in precise terms

About 84% of the reported decline from fall 2000 to fall 2025 happened before EFAs began. That is a valid correction to claims that EFAs caused the whole decline. The separate 3% ratio does not measure the program’s causal effect: it compares reported transfers with a much longer net enrollment change.

[23] [22] [45] [47] [8]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The 84% timing point is sound: the program cannot cause a pre-program decline. But dividing five years of reported transfers by a net enrollment change over 25 years does not estimate causation. Neither a switcher count nor a small statewide share rules out state-budget costs or district pressure.[23] [22] [45] [47] [8]

Check the calculation

1,437 ÷ 48,139 = 2.985%. That denominator includes two decades before EFAs. Even 1,437 ÷ (168,620−160,322) = 17.32%, using reported fall 2021/2025 totals, would not be a causal estimate: it mixes gross transfers with net population change and needs unique-student and coverage checks.

[23] [22] [45] [47] [8]

Source trail

23EFAsIncomplete without rollover rule

Unused EFA money does not return every year.

The statute allows unused funds to roll over quarter to quarter and year to year while the EFA remains in force. It provides for reversion when the account ends.[21]

Claim being checked

Unused money / return to state

The site says unused EFA money goes back to the Education Trust Fund.

View related original graphic ↗
What the primary record shows

RSA 194-F:2 VII provides both rollover and reversion; IV requires vendor refunds to be credited back to the EFA. Funds remain restricted to authorized education uses.[21]

What this means, in precise terms

NH law permits unused EFA funds to roll over while an account remains in force. It is misleading to imply all unspent awards automatically return to the state at year end.

[21]
Calculation, qualifications & source trail A

Where the framing can mislead—or the evidence stops

A reader could mistake “unused money returns” for annual return of unspent balances. Annual awards, annual expenses, account balances and amounts reverted are separate fiscal measures.[21]

Check the calculation

Awarded funds − eligible spending + refunds = change in account balance, subject to transfers and closure/reversion. An award total is not necessarily the amount spent by families that fiscal year.

[21]

Source trail

24ResultsSupported with context

Grade 8 math deterioration is real. Causation is unresolved.

The NAEP snapshot confirms the decline and the 31% below-Basic figure. NH still scored above the national public-school average in 2024.[24] [28]

Claim being checked

286 → 280; 21% → 31%

The site reports grade 8 math scores down since 2003, with nearly one-third below Basic in 2024.

View related original graphic ↗
What the primary record shows

NH grade 8 math: 286 in 2003, 280 in 2024; below Basic: 21%→31%. NCES marks the long-run mean decline and Basic-level change significant. NH 2024 mean 280 exceeds U.S. public 272.[24] [28]

What this means, in precise terms

NAEP confirms NH grade 8 math declined from 2003 to 2024, with 31% below Basic in 2024. NH remained above the U.S. public-school mean, and the trend alone does not identify the cause.

[24] [28]
Calculation, qualifications & source trail A

Where the framing can mislead—or the evidence stops

2024 vs. 2022 mean change (279→280 rounded) is not statistically significant. NAEP Proficient is a demanding performance standard, not a synonym for grade-level proficiency. Trend comparisons are descriptive; they do not isolate spending, staffing, pandemic or demographic effects.[24] [28]

Check the calculation

Below Basic increased 10 percentage points, or 47.62% relative to 21%; those are different measures. In 2024, 69% were at/above Basic and 32% at/above Proficient (rounded categories may not sum exactly).

[24] [28]

Source trail

25ResultsRounded arithmetic supported; inference limited

The score sum checks out; it is not a NAEP index.

The rounded score sum is reproducible from primary NCES tables. It is a home-built composite, not an official NAEP index or evidence that spending caused lower scores.[24] [26]

Claim being checked

−21; −7; 3×

The site adds four grade/subject score changes and says NH’s decline was three times the national decline.

View related original graphic ↗
What the primary record shows

For 2003→2024, rounded NH changes are: grade 4 reading 228→221 (−7), grade 8 reading 271→264 (−7), grade 4 math 243→242 (−1), grade 8 math 286→280 (−6). U.S. public changes are 216→214 (−2), 261→257 (−4), 234→237 (+3), 276→272 (−4). All four NH means remain above national public means in 2024.[24] [26] [25] [27] [36] [37] [38] [39]

What this means, in precise terms

The site’s rounded NAEP sum can be reproduced: −21 points for NH versus −7 nationally across four tests. That composite is not an official NAEP index and cannot establish that spending caused the declines.

[24] [26] [25] [27] [36] [37] [38] [39]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

Separate tests have separate scales and uncertainties. Equal weighting of grades and subjects is a choice, not a supplied measure of learning lost. Grade 4 math’s long-run mean change is not significant. The grade 4 reading snapshot uses 1998; the 2003 figure comes from Digest Table 221.40. No supplied test establishes a significant threefold difference in aggregate decline.[24] [26] [25] [27] [36] [37] [38] [39]

Check the calculation

NH rounded sum: −7−7−1−6 = −21. U.S.: −2−4+3−4 = −7. Ratio of absolute sums: 21÷7 = 3. Digest Table 222.60 displays NH 2003 grade 8 math as 286.2; rounded to whole points consistently with the 2024 snapshots, it is 286. Using inconsistent precision gives a slightly different sum.

[24] [26] [25] [27] [36] [37] [38] [39]

Source trail

26ResultsCalculation reproduced; interpretation needs limits

The 73% / four-point comparison is real—and easy to overread.

The original NHED cost and assessment files reproduce 72.88% higher spending and a 4.01-percentage-point proficiency difference. That is a comparison of equally weighted districts, not a causal estimate.[5] [71]

Claim being checked

73%; 54%; 50%; 44%; 55%

The site says the highest-spending district quarter spends 73% more for four points higher proficiency; it also reports 2026 statewide math and reading results. The related scorecard also reports 33% versus 26% economically disadvantaged students among districts with a reported figure. This is the dashboard’s 2025 economic-disadvantage indicator.

View related original graphic ↗
What the primary record shows

The 158 matched districts have exact published math and reading percentages. The lowest-cost 39 average $19,930.01 per pupil and 50.141% combined proficiency; the highest-cost 39 average $34,455.44 and 54.154%. All included values match the site. NHED’s separate 2026 statewide release reports math proficiency 44% and reading/writing 55%. The separate 33% / 26% disadvantage comparison also reproduces from NHED’s live iExplore dashboard, after applying the original footnote’s available-data rule.[5] [71] [72] [29] [47] [78] [73] [74]

What this means, in precise terms

The site’s arithmetic checks out, but “73% more spending for only four points” is not a measured return on investment. It compares different districts with different sizes and grade ranges. The four-point gap gives every district equal weight; using the published tested-count bands instead gives a gap of roughly ten points. That sensitivity is a reason to explain the comparison carefully, not to claim either result proves spending does or does not work.

[5] [71] [72] [29] [47] [78] [73] [74]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

This averages proficiency percentages, not raw scores: each subject and each district receives equal weight. Small districts therefore count as much as large ones. NHED cautions that costs should be compared across the same grade ranges, yet the groups mix ranges; 23 low-cost districts versus eight high-cost districts have reported tested grades 3–8 and 11. Needs, scale, grade mix and prior outcomes are not controlled. One district with >90% censored rates is excluded; including its permitted interval changes the original-style gap to >4.94 and ≤5.19 points. A single-year comparison cannot show what an additional dollar causes.[5] [71] [72] [29] [47] [78] [73] [74]

Check the calculation

Join FY2025 operating cost to school-year 2024–25 district / All grades / All students / Regular-denominator math and reading results. Of 162 numeric-cost matches, remove three suppressed results and one >90% censored result. Sort the remaining 158 by cost; compare bottom and top floor(158/4)=39. District proficiency = (math% + reading%)/2. Cost premium = (34,455.444359 ÷ 19,930.011538 − 1) × 100 = 72.8822%. Proficiency difference = 54.153846 − 50.141026 = 4.01282 percentage points. Across all 158, Pearson r ≈ 0.1033: a weak positive descriptive association, with no causal identification.

[5] [71] [72] [29] [47] [78] [73] [74]
The published district-weighted calculation, independently reproduced
MeasureLowest-cost 39Highest-cost 39Difference
Operating cost per pupil$19,930.01$34,455.44+72.88%
Mean math proficiency45.051%48.615%+3.564 points
Mean reading proficiency55.231%59.692%+4.462 points
Equal-subject, equal-district mean50.141%54.154%+4.013 points
Tested-count-weighted midpoint illustration45.501%55.661%+10.160 points

Why the weighting result is approximate: the primary file reports tested counts in bands. Holding its rounded proficiency percentages fixed and allowing counts anywhere inside those bands gives a conservative weighted gap of 9.85–10.47 percentage points. This pools math and reading assessment observations; it is not the percentage of unique children proficient in both. Rate-rounding uncertainty is additional.

Original disadvantage comparison reproduced • 2025 NHED iExplore
MeasureLowest-cost 39Highest-cost 39
Districts with numeric reported percentages2927
Excluded censored / suppressed values1012
Sum of reported district percentages753.71878.36
Mean among numeric reports25.9900%32.5319%
Rounded result used by original site26%33%

The original caveat deserves credit: the site says “among districts with a reported figure.” Excluding <10% and *N entries reproduces its numbers. These are available-case district means, not the disadvantaged share of all students in either full 39-district group. Missingness is systematic: all ten exclusions in the low-cost group are <10%; the high-cost group has three <10% entries and nine *N entries. Averages over different available subsets need that caveat.

Why the earlier check missed: October meal-eligibility reports and tested-student SES counts are different measures. The directly linked iExplore indicator resolves this claim; the alternative datasets should not be used to declare it false.

Source trail

27EFAsGrant verified; cost comparison incomplete

Charters are public schools; a grant is not total cost.

The FY2025 $9,180 grant total is verified from the original agency rate history. Comparing that grant with district all-source spending still leaves a material difference in cost definitions.[33] [43]

Claim being checked

$4,182 + $4,998 = $9,180

The site correctly describes charter schools as public schools and contrasts their state grant with district operating cost.

View related original graphic ↗
What the primary record shows

The recovered NHED packet’s rate-history table confirms FY2025 base aid $4,182 and the non-VLACS additional grant $4,998, totaling $9,180 before differentiated amounts. The packet also specifies that special-education differentiated aid goes to the district of financial responsibility; FY2025 enrollment/ADM figures are estimates.[33] [43] [5]

What this means, in precise terms

The FY2025 $9,180 non-virtual charter grant is verified, and charters are public schools. Comparing that state grant with district schools’ all-source average expenditure is incomplete: it does not measure the full cost of equivalent services or the effect of a transfer.

[33] [43] [5]
Calculation, qualifications & source trail A + B

Where the framing can mislead—or the evidence stops

The online packet has been refreshed and includes FY2026 rates, so the FY2025 history row must be selected explicitly. A charter grant is not all-source expenditure. District-provided services, other charter revenues, grade mix and different obligations must be included before claiming the full cost of an equivalent education.[33] [43] [5]

Check the calculation

For a comparable cost study, align FY2025, actual expenditure, pupil membership, grade mix and services; include all public revenues and any district-provided services without double counting. A state grant is a financing input, not audited total cost.

[33] [43] [5]

Source trail

28TaxesSupported within its legal scope

A rejected SB2 budget does not automatically close schools.

RSA 40:13 supports the basic point for jurisdictions that adopted the official-ballot procedure. That does not guarantee unchanged programs or staffing.[32]

Claim being checked

default budget

The site says a rejected proposed operating budget activates the default budget in SB2 districts.

View related original graphic ↗
What the primary record shows

The default is based on prior-year appropriations with statutory adjustments for existing obligations and specified reductions; the law also provides a revised-budget meeting mechanism.[32]

What this means, in precise terms

In districts using SB2, rejecting the proposed budget generally activates the statutory default budget. That avoids an automatic shutdown, but it does not guarantee unchanged services.

[32]
Calculation, qualifications & source trail A

Where the framing can mislead—or the evidence stops

It is not necessarily identical to last year’s spending, not applicable to every governance system, and not a universal guarantee that a lower budget has no educational effect. Contractual obligations can leave discretionary services bearing more adjustment.[32]

Check the calculation

Compare the actual proposed and certified default budgets line by line. Identify required obligations, removed one-time items, eliminated positions and projected revenues. No statewide dollar reduction follows just from voting “no.”

[32]

Source trail

Change one assumption.
See what changes.

These are illustrative models, not district forecasts. They show why a smaller grant and a smaller tax levy do not automatically reveal which costs are avoided or which services change.

EFA / combined public cost

Hold the year and service comparison constant. Enter the grant and the public costs actually avoided. Transfers between public budgets cancel when calculating combined taxpayer cost.

Illustration: $5,000 − $2,000 = $3,000 additional combined public cost.

If the student would otherwise remain privately educated without public subsidy, avoided public-school cost can be zero. A public-school switcher may have different costs. This simple model excludes administration and other consequential costs.

Local levy pressure from a transfer

Illustration: $5,000 − $2,000 − $0 = $3,000 more local revenue needed to preserve the remaining services.

This holds other revenues and services constant. Aid formulas, timing, enrollment weights and statutory transition provisions must be checked for the relevant fiscal year. A higher local requirement does not automatically produce that exact change in an individual bill.

The full policy timeline

Actions, actors and controversies are dated. Laws, court holdings, estimates and study proposals are kept distinct.

  1. 1993 & 1997

    Constitutional duty and tax uniformity

    Governors Steve Merrill (R, 1993) and Jeanne Shaheen (D, 1997); Republican-controlled legislatures in these periods. These are court decisions, not bills signed by the governors.

    The school-funding decisions established a state duty to fund adequate educational opportunity and constraints on disproportionate taxation. The 2025 ConVal opinion restates those holdings. Controversy: scope of state responsibility, local control and judicial authority.[11]

  2. 1999

    A statewide tax and broader state revenues

    Governor Jeanne Shaheen (D); Republican House and Senate.

    HB117, Chapter 17, established a uniform education property tax alongside other revenue measures. Controversy: how to spread a statewide duty across unequal property-tax bases.[14]

  3. 2005

    A fixed $363 million SWEPT target

    Governor John Lynch (D); Republican House and Senate. HB616 became law without his signature.

    Chapter 257 set the nominal target and reworked the formula. The target is not an indexed guarantee that state revenues will keep pace with school costs. Controversy: tax capacity, redistribution and constitutional adequacy.[15] [13]

  4. 2009

    State retirement share reduced

    Governor John Lynch (D); Democratic House and Senate.

    HB2, Chapter 144:52, reduced the state share to 30% in FY2010 and 25% in FY2011, with a scheduled return to 35%. Controversy: state budget relief versus local employer obligations.[16] [19]

  5. 2011 → FY2013

    Retirement subsidy ended; excess SWEPT retained

    Governor John Lynch (D); Republican House and Senate. HB2 became law without his signature; the official HB337 record identifies a signed law.

    HB2 replaced the retirement payment with $3.5m in FY2012 and ended it in FY2013. The legislature cited expected reform savings. The 2025 Rand opinion records excess retention since 2011 and discusses HB337. Controversy: local cost shifting and whether equal rates produce fair funding.[17] [18] [19] [12]

  6. 2021

    Education Freedom Accounts created

    Governor Chris Sununu (R); Republican House and Senate.

    HB2, Chapter 91:431, created RSA 194-F. Controversy: family choice versus new public subsidies, oversight, access and effects on district resources. This is the original enactment; subsequent eligibility rules have changed.[23] [21]

  7. 2022 → FY2023

    One-time retirement assistance

    Governor Chris Sununu (R); Republican House and Senate.

    HB1221, Chapter 189, funded an amount equal to 7.5% of eligible employer contributions; LBA records $27.1m spent in FY2023. One-time assistance did not restore the historical permanent 35% arrangement.[20] [19]

  8. 2023

    Base and differentiated aid reset

    Governor Chris Sununu (R); Republican House and Senate.

    Chapter 79:150 set a $4,100 base with differentiated aid, subject to adjustment. The FY2025 agency explanation gives $4,182. Debate: a legislative rate versus evidence of the cost of the statutory educational opportunity.[9] [10]

  9. June 10, 2025

    EFA expansion enacted

    Governor Kelly Ayotte (R); Republican House and Senate.

    SB295, Chapter 75, expanded EFA eligibility. The official docket records signature on June 10 and a House final vote of 190–178. Debate concerned broader family choice, state costs and program growth. Historical costs before expansion cannot establish the later program’s budget impact; check the enacted chapter and current rules for the applicable year.[40] [21]

  10. June 10, 2025

    Rand: excess retention upheld; negative rates rejected

    Governor Kelly Ayotte (R); Republican House and Senate. The Supreme Court decided the case.

    The majority treated excess retention as a spending decision outside the asserted tax-uniformity violation. It affirmed the negative-rate violation. The dissent disagreed on retention. A legal ruling does not settle policy fairness.[12]

  11. July 1, 2025

    ConVal: inadequacy affirmed; immediate payment reversed

    Governor Kelly Ayotte (R); Republican House and Senate. The Supreme Court decided the case.

    The Court affirmed the constitutional defect and conservative minimum guidepost above $7,356.01 under the existing framework; it reversed the directive for immediate payments. Responsibility to remedy the defect remained with elected branches.[11]

  12. November 1, 2025

    SAU committee proposes a county model

    Governor Kelly Ayotte (R); Republican House and Senate. Legislative study under SB57, Chapter 220.

    The report favors concentrating business functions and cites the 1993 savings model. Educator testimony raises service, geography and transition concerns. A study recommendation is not proof of enacted statewide reorganization or realized savings.[3]

  13. July 15, 2026

    HB1300 makes the local cap question mandatory

    Governor Kelly Ayotte (R); Republican-controlled House and Senate.

    The final conference report passed House 183–170 and Senate 15–8. Chapter 324 places cap questions on the November 2026 and 2028 ballots. This is a new local-revenue and administrative-budget policy, not an order implementing ConVal. Supporters emphasize taxpayer protection; the separate levy formula is why service effects remain a material question.[55] [56] [58]

  14. September 28, 2026

    Ballot petition denied without a merits explanation

    NH Supreme Court; Petition of David Bates & a., No. 2026-0612.

    The one-page order denies original jurisdiction under Rule 11. It does not supply a substantive analysis of the ballot’s “do not affect classroom instruction” sentence or settle every future implementation dispute.[57]

  15. November 3, 2026 → FY2028

    Voters decide local adoption

    Scheduled general election; outcomes unknown as of October 24, 2026.

    Three-fifths approval is required under the law’s district and SAU tally provisions. A cap approved in November 2026 begins with FY2028; a cap approved in November 2028 begins with FY2030. Overrides and certified bonded-capital exclusions matter. The statutory scheme expires January 1, 2032.[55]

  16. 2026 • currency check

    Do not freeze a moving policy debate in 2025

    Governor Kelly Ayotte (R); Republican House and Senate.

    The live statutes show 2026 amendments to adequacy and EFA law. This edition verifies the specified 2025 opinions, not the complete subsequent litigation docket or all 2026 fiscal implementation. Before using this for a current legal or budget decision, check later orders, enacted chapters and the applicable year’s NHED award schedule.[9] [21]

Verification record and methods

The original figures are tested against the original data and definitions. A different, nearby calculation does not establish that a published number is wrong.

How the last gaps were resolved

  • Recent financing: the January 7, 2025 NHED release, the report’s fall-enrollment counts and the “other local revenue” category reproduce all six dollar amounts and the rounded 19% / 29% / 126% increases. Exact components →
  • Teacher pay and spending: historical PCE observations reproduce 7.61% and 44.57%, rounding to 8% and 45%. The same inflation factor reproduces the report’s printed $2,807,974,459 starting-spending baseline to the nearest dollar. NCES’s total-expenditure definition resolves the original $1,723,024,544 nominal baseline. The 45% growth calculation changes definition by adding debt principal only at its 2024 endpoint; removing that addition gives 41.76% under the same inflation convention. Full reconstruction →
  • District results: all 158 included cost/proficiency records match the primary data; the 73% / four-point result is reproduced. The live 2025 iExplore disadvantage indicator also reproduces 33% / 26% among numerical reports. Suppressed and censored entries are excluded exactly as the original footnote describes. Data, exclusions and weighting →
  • Historical financing: a 170-district panel, statewide pupil counts, historical January PCE and export precision reproduce every original revenue endpoint. Aligning the pupil count with the same district panel changes −3% / +135% to about −0.27% / +142.22%. Original source and method →

Corrections to this audit: earlier editions used a newer financing release, a narrower NCES expenditure field, fiscal-year inflation averages and different disadvantage datasets. Those are not interchangeable with the original methods. This edition gives the original calculations proper credit and makes any separate sensitivity comparison explicit.

  • Source access: original records were recovered directly, from Internet Archive, from a government-hosted copy, or through a text retrieval proxy. The live public NHED dashboard was retrieved through its new official embedding page. These are access routes to the agency record, not additional independent endorsements. Source dates, versions and links are listed below.
  • Years and pupils: reproduce the cited method first; then disclose its limits. Fall enrollment, finance-adjusted membership, ADM-A and tested students have different populations. The original FY2024 NHED release excludes a nonreporting district. Historical comparisons are not automatically fixed-district panels.
  • Inflation: the original reports’ January-to-January PCE factors are reconstructed from historical ALFRED data vintages and exact chart/workbook matches. School-year salary sensitivity uses July–June PCE or CPI-U averages. Historical NHED charts use their own stated constant-dollar basis, and the property-tax comparison uses July-to-July CPI. Index rebasing alone cancels in a ratio; data revisions and different months can change the result.
  • Scores: equal-district, equal-subject means differ from pooled assessment observations. Reported count bands, rounded proficiency rates, grade coverage, suppression, censored values and available-case means are disclosed. None of these descriptive comparisons identifies the causal effect of spending.
  • Evidence order: original agency records, laws and court-authored decisions control factual checks. An author’s workbook is evidence of the author’s method; its figures still need checking against primary data. NEA is the original salary-survey producer used by NCES and also an advocacy organization. Neither an organization’s advocacy nor its official status makes an untested causal inference conclusive.
  • History and motives: governors, legislative majorities, votes, enactments and court holdings are sourced. Chronology does not establish why a website was created. The court cases and local ballot are separate legal events; the brief 2026 procedural order is not treated as a merits judgment on classroom effects.
  • Scope: major statewide main/share-page claims captured October 24, 2026. Videos and every municipal output are outside this audit; no town-specific examples are used. The editorial scores assess accuracy, presentation and balance separately and do not claim to measure the authors’ intent.
  • Offline use: this file has no external assets, tracking or dependencies. Source links need internet. Search and filters work locally; print includes all claims and calculations. Desktop, mobile, light/dark, search, controls and print were checked. This is not a full assistive-technology certification.

Sources you can inspect

Numbered citations beside factual paragraphs link here. Each record identifies the organization, period and what was inspected. Retrieval date: October 24, 2026. Stable archive copies are linked where available; live agency files can be revised.

  1. [1] Census • FY2024 summary workbook, Tables 1, 8 & 19 ↗

    Final data page revised May 7, 2026. Downloaded and cells read independently.

    Retrieved / checked October 24, 2026.
  2. [2] NCES • Digest 2023, Table 213.50 ↗

    Fall 2012–2022; table downloaded and NH/U.S. rows read.

    Retrieved / checked October 24, 2026.
  3. [3] NH Legislature • SAU study final report ↗

    November 1, 2025; pp. 2–6 and appended 1993 study. Read independently; testimony and estimates distinguished from outcomes.

    Retrieved / checked October 24, 2026.
  4. [4] RSA 194-C:4 • Superintendent services ↗

    Current online text checked October 24, 2026; amendment history through 2010.

    Retrieved / checked October 24, 2026.
  5. [5] NHED • Cost per pupil by district, FY2025 ↗

    Issued January 8, 2026; ADM-A as of August 12, 2025. Original NHED PDF recovered from Internet Archive capture February 4, 2026. Independently extracted; statewide row $22,699.85; introductory broader measure $27,689.90.

    Retrieved / checked October 24, 2026.
  6. [6] NHED • Cost per pupil increases by district, 2000–2023 ↗

    Original four-page agency chart; Internet Archive capture June 9, 2024. Pages rendered and OCR rows checked. $11,192→$20,323 statewide, in 2023 dollars. 158 district rows contain both amounts; all 158 increase. Four repeated statewide rows excluded; two OCR-missed rows checked visually. The site’s 158-record selection has now been reconciled to the original chart.

    Retrieved / checked October 24, 2026.
  7. [7] NHED • Enrollment report directory ↗

    Directory inspected; the original fall 2011, 2020, 2023 and 2025 reports and final fall 2025 SAU report were subsequently retrieved through their PrintReport endpoints. Individual source entries below identify those reads.

    Retrieved / checked October 24, 2026.
  8. [8] NCES • Digest 2013, Table 203.20 ↗

    Historical state public-school enrollment. Downloaded table independently confirms NH fall 2000 enrollment of 208,461.

    Retrieved / checked October 24, 2026.
  9. [9] RSA 198:40-a • Base and differentiated aid ↗

    Current text checked October 24, 2026; includes 2026, 19:3, effective May 26, 2026. Statutory schedule is subject to adjustments.

    Retrieved / checked October 24, 2026.
  10. [10] NHED • FY2025 adequacy explanation ↗

    September 2024 explanation, all three pages independently recovered through a text retrieval proxy. Confirms FY2025 base $4,182 per ADM, prior-year resident ADM, differentiated rates, SWEPT crediting, extraordinary-needs aid, hold-harmless provisions and payment schedule. Historical formula explanation; later amendments are not inferred from this document.

    Retrieved / checked October 24, 2026.
  11. [11] NH Supreme Court • ConVal, 2025 N.H. 29 ↗

    July 1, 2025. Court-authored opinion, read in full text via an ACLU-hosted copy; see ¶¶ 2, 62–78. Hosting organization is an advocate; source is the judicial opinion.

    Retrieved / checked October 24, 2026.
  12. [12] NH Supreme Court • Rand, 2025 N.H. 27 ↗

    June 10, 2025. Court-authored opinion via advocate-hosted copy; ¶¶ 1–2, 17 and negative-rate analysis; includes dissent.

    Retrieved / checked October 24, 2026.
  13. [13] RSA 76:3 • State education property tax ↗

    Current online text checked October 24, 2026; $363 million target, dating from 2005.

    Retrieved / checked October 24, 2026.
  14. [14] 1999 HB117 • Chapter 17 ↗

    Enacted 1999; creation of uniform education property tax and other revenue measures.

    Retrieved / checked October 24, 2026.
  15. [15] 2005 HB616 • Chapter 257 ↗

    Became law July 16, 2005 without Governor John Lynch’s signature; $363 million tax target.

    Retrieved / checked October 24, 2026.
  16. [16] 2009 HB2 • Chapter 144 ↗

    Enacted 2009; §52 changes state retirement contribution shares.

    Retrieved / checked October 24, 2026.
  17. [17] 2011 HB2 • Chapter 224 ↗

    Enacted 2011 without Governor John Lynch’s signature; §§191–192 retirement changes. Includes suspension of new school-building aid for the biennium, with waiver provision.

    Retrieved / checked October 24, 2026.
  18. [18] General Court • 2011 HB2 status ↗

    Official docket identifies LAW WITHOUT SIGNATURE, Chapter 224.

    Retrieved / checked October 24, 2026.
  19. [19] Legislative Budget Assistant • State Aid FY2025 ↗

    Retirement Normal Contribution, p. 105; historical shares, FY2012 $3.5m and FY2023 $27.1m one-time aid verified in agency text.

    Retrieved / checked October 24, 2026.
  20. [20] General Court • 2022 HB1221 status ↗

    Chapter 189, signed by governor; one-time retirement assistance.

    Retrieved / checked October 24, 2026.
  21. [21] RSA 194-F:2 • EFA funding and use ↗

    Current text checked October 24, 2026; history includes 2025 and 2026 amendments. Base plus differentiated aid, permitted uses and rollover.

    Retrieved / checked October 24, 2026.
  22. [22] NHED • November 2025 EFA data fact sheet ↗

    Original 14-page NHED PDF; Internet Archive capture December 3, 2025. Tables 1–3, 9–12 independently read. Year labels are ending school years. Table 3 is projected annual awards, not audited family expenditures; 2025/2026 transfer figures are preliminary.

    Retrieved / checked October 24, 2026.
  23. [23] 2021 HB2 • EFA authorizing law ↗

    Chapter 91:431 created RSA 194-F. Historical enactment, not a statement of today’s eligibility rules.

    Retrieved / checked October 24, 2026.
  24. [24] NCES • 2024 NAEP grade 8 mathematics ↗

    Published January 2025; primary PDF downloaded and read, 2003/2022/2024.

    Retrieved / checked October 24, 2026.
  25. [25] NCES • 2024 NAEP grade 4 mathematics ↗

    Published January 2025; primary PDF downloaded and read.

    Retrieved / checked October 24, 2026.
  26. [26] NCES • 2024 NAEP grade 8 reading ↗

    Published January 2025; primary PDF downloaded and read.

    Retrieved / checked October 24, 2026.
  27. [27] NCES • 2024 NAEP grade 4 reading ↗

    Published January 2025; primary PDF downloaded and read. Snapshot baseline is 1998, not 2003.

    Retrieved / checked October 24, 2026.
  28. [28] NCES • Interpreting NAEP mathematics ↗

    Official explanation of scale scores and achievement levels. Read October 24, 2026.

    Retrieved / checked October 24, 2026.
  29. [29] NHED • 2026 assessment results, September 2, 2026 ↗

    Original agency release recovered from Internet Archive capture September 22, 2026. Confirms 44% mathematics proficiency in 2026, up from 42% in 2025, and 55% reading/writing, unchanged. These are tested-student proficiency rates, not all children’s grade-level status.

    Retrieved / checked October 24, 2026.
  30. [30] NH DRA • 2025 municipal tax rates, completed January 23, 2026 ↗

    Original DRA statewide report on a municipal government server. Its bytes match the January 30, 2026 Internet Archive capture of the DRA original. All 260 rows extracted, including 234 incorporated municipalities, a separate district line and unincorporated places. Rates multiplied by applicable values; gross components distinguished from commitments net of adjustments.

    Retrieved / checked October 24, 2026.
  31. [31] NHED • Financial reports ↗

    Primary directory inspected. Original FY2020, FY2024 and FY2025 tables were retrieved and read; independent historical benchmarks now use NCES state-finance files with matched membership. Different agency definitions and revisions are labeled, not blended.

    Retrieved / checked October 24, 2026.
  32. [32] RSA 40:13 • Official ballot and default budget ↗

    Current statute read October 24, 2026. Applies to adopting subdivisions; see IX–X and related default-budget provisions.

    Retrieved / checked October 24, 2026.
  33. [33] NHED / House Finance • Charter funding orientation and rate history ↗

    Original five-page agency packet retrieved via text proxy; online file last-modified July 2026, internal rate-history table January 2025. FY2025 base $4,182 + non-VLACS grant $4,998 = $9,180. FY2026 rates differ. FY2025 charter count is estimated/enrollment-based, not final ADM.

    Retrieved / checked October 24, 2026.
  34. [34] Scafidi / Josiah Bartlett Center • 2001–2019 ↗

    March 2023 policy analysis. Published district workbook recovered; original revenue panel and January PCE adjustment independently reproduced. Its revenue panel and statewide pupil denominators have different coverage. Advocacy is not treated as evidence against the numerical inputs; causal claims are assessed separately.

    Retrieved / checked October 24, 2026.
  35. [35] Scafidi / Josiah Bartlett Center • 2020–2024 ↗

    April 2025 policy analysis. Original NHED snapshot, all six 2020/2024 funding endpoints, salary endpoints, January PCE factor and rounded 8%/45% arithmetic reproduced. The long-run expenditure comparison changes definition by adding debt principal only to its later endpoint. Source notes disclose the omitted FY2024 reporting district.

    Retrieved / checked October 24, 2026.
  36. [36] NCES • Digest 2017, Table 221.40 • grade 4 reading ↗

    Prepared March 2018; downloaded and 2003 NH/U.S. public-school baseline cells inspected independently. Whole-point rounding used for the composite.

    Retrieved / checked October 24, 2026.
  37. [37] NCES • Digest 2017, Table 221.60 • grade 8 reading ↗

    Prepared March 2018; downloaded and 2003 NH/U.S. public-school baseline cells inspected independently. Whole-point rounding used for the composite.

    Retrieved / checked October 24, 2026.
  38. [38] NCES • Digest 2017, Table 222.50 • grade 4 math ↗

    Prepared March 2018; downloaded and 2003 NH/U.S. public-school baseline cells inspected independently. Whole-point rounding used for the composite.

    Retrieved / checked October 24, 2026.
  39. [39] NCES • Digest 2017, Table 222.60 • grade 8 math ↗

    Prepared March 2018; downloaded and 2003 NH/U.S. public-school baseline cells inspected independently. Whole-point rounding used for the composite.

    Retrieved / checked October 24, 2026.
  40. [40] General Court • SB295, Chapter 75, 2025 ↗

    Official docket: signed June 10, 2025; House final vote 190–178. Expansion legislation; consult the chapter and current RSA 194-F for eligibility and cap details.

    Retrieved / checked October 24, 2026.
  41. [41] NHED • State average cost and total expenditures, FY2025 ↗

    January 8, 2026; Internet Archive capture February 19, 2026. Original one-page agency table read. Part A total $4,157,349,112; ADM-A 149,385.66; operating $22,699.85; elementary/secondary $27,689.90; all listed programs $27,829.65 per ADM-A. Bond principal excluded to prevent double counting.

    Retrieved / checked October 24, 2026.
  42. [42] NH DRA • Archived original 2025 rate report ↗

    Original DRA URL preserved in January 30, 2026 capture. Same 270,660-byte PDF as the municipal mirror. Report completed January 23, 2026. Used for independent statewide aggregation, not individual bill predictions.

    Retrieved / checked October 24, 2026.
  43. [43] NHED • FY2026 adequate education aid explained ↗

    November 2024 explanation; original PDF captured January 11, 2025. Independently read: FY2026 base $4,265.64, differentiated rates, prior-year ADM, SWEPT crediting and estimate/payment schedule. Historical estimate-stage explanation, not a certification of later amendments or final awards.

    Retrieved / checked October 24, 2026.
  44. [44] NHED • District fall enrollment, October 2011 ↗

    Original live agency PrintReport output retrieved October 24, 2026 through a text retrieval proxy after direct access was blocked. Structured report rows read, duplicate print representations removed, district IDs retained. SchoolYear parameter is the ending year, so 2012 means fall 2011.

    Retrieved / checked October 24, 2026.
  45. [45] NHED • District fall enrollment, October 2020 ↗

    Original live agency PrintReport output retrieved October 24, 2026 through a text retrieval proxy after direct access was blocked. Structured report rows read, duplicate print representations removed, district IDs retained. SchoolYear parameter is the ending year, so 2021 means fall 2020.

    Retrieved / checked October 24, 2026.
  46. [46] NHED • District fall enrollment, October 2023 ↗

    Original live agency PrintReport output retrieved October 24, 2026 through a text retrieval proxy after direct access was blocked. Structured report rows read, duplicate print representations removed, district IDs retained. SchoolYear parameter is the ending year, so 2024 means fall 2023. Traditional district total 155,253. Two rows carrying financial district ID 388 total 1,173 pupils (1,019 + 154); excluding both gives 154,080, exactly the original FY2024 policy report denominator. Exclusion by district identifier, not just displayed name, is required.

    Retrieved / checked October 24, 2026.
  47. [47] NHED • District fall enrollment, October 2025 ↗

    Original live agency PrintReport output retrieved October 24, 2026 through a text retrieval proxy after direct access was blocked. Structured report rows read, duplicate print representations removed, district IDs retained. SchoolYear parameter is the ending year, so 2026 means fall 2025.

    Retrieved / checked October 24, 2026.
  48. [48] NHED • SAU enrollments, October 1, 2025 ↗

    Report ID 6; Final; report date February 3, 2026; run October 24, 2026. Original agency HTML retrieved via text proxy. 203 unique district rows sum to 160,322. For SAU numbers below 200: 101 unique SAUs, 50 with summed enrollment below 1,000. Academies/charters excluded from that SAU count, not silently merged.

    Retrieved / checked October 24, 2026.
  49. [49] NCES • Digest 2015, Table 213.50 ↗

    Primary table: fall 2012 NH staff 31,734 FTE, pupils 188,974. Independently downloaded; compared with fall 2022 counts from Digest 2023. Counts, not rounded pupil/staff ratios, used in the calculation.

    Retrieved / checked October 24, 2026.
  50. [50] NCES • Digest 2003, Table 81 • fall 2000 staffing ↗

    Primary NCES table independently downloaded. NH FTE: teachers 14,341; aides 5,056; district officials/administrators 439; instructional coordinators 173; student support 542. Fall 2000 corresponds to the 2000–01 school year.

    Retrieved / checked October 24, 2026.
  51. [51] NCES • Digest 2020, Table 213.20 • fall 2018 staffing ↗

    Primary NCES table independently downloaded. NH FTE: teachers 14,644; aides 7,061; district officials/administrators 690; coordinators 278; student support 767. Fall 2018 corresponds to 2018–19; categories/versions can differ from earlier releases.

    Retrieved / checked October 24, 2026.
  52. [52] NHED • Valuations, assessments and rates, 2025–26 ↗

    October 2026 revision of 2025 valuations. Original agency document text retrieved via proxy. Published totals: local school assessment $2,790,227,910; other local $1,952,547,089; state school $363,492,715, plus $24,121 from omitted unincorporated places. Reference-only cooperative rows must not be summed again.

    Retrieved / checked October 24, 2026.
  53. [53] NHED • District revenue and expenditures, FY2020 ↗

    December 21, 2020. Original agency one-page table retrieved via text proxy. Net revenue $3,355,453,164; state sources $1,055,087,086; local tax $2,099,195,588; federal $161,745,104. Interdistrict revenue/expenditure adjustments disclosed.

    Retrieved / checked October 24, 2026.
  54. [54] NHED • District revenue and expenditures, FY2024 ↗

    Revised January 6, 2026. Original agency one-page table retrieved via text proxy. Net revenue $4,041,009,041; state sources $1,158,775,334; local tax $2,461,935,991; federal $339,939,204. Different coverage/revisions from older snapshots must not be blended.

    Retrieved / checked October 24, 2026.
  55. [55] 2026 HB1300 • Chapter 324 • enacted law ↗

    Approved July 15, 2026; substantive provisions effective September 1, 2026. Eight-page final chapter independently read. RSA 32:5-i V (ballot text), VI (adoption), VIII–IX (separate caps), X (overrides), XII (FY2028/FY2030), XVI (bond exclusions), XVII (2032 sunset).

    Retrieved / checked October 24, 2026.
  56. [56] General Court • HB1300 legislative history and votes ↗

    Introduced January 7, 2026. Conference report adopted June 4: House 183–170; Senate 15–8. Signed by Governor Kelly Ayotte July 15, 2026, Chapter 324. Docket inspected independently; records enactment rather than a court remedy.

    Retrieved / checked October 24, 2026.
  57. [57] NH Supreme Court • Petition of David Bates & a., 2026-0612 ↗

    September 28, 2026, one-page primary order retrieved via text proxy. It denies the petition for original jurisdiction, citing Rule 11. It gives no substantive explanation and does not find that classroom services cannot be affected.

    Retrieved / checked October 24, 2026.
  58. [58] Governor • Statement on signing HB1300 ↗

    July 15, 2026 signing announcement, read from the original government page via text proxy. Primary evidence of the administration’s stated taxpayer-protection rationale; advocacy for the policy, not proof of its eventual effects.

    Retrieved / checked October 24, 2026.
  59. [59] Secretary of State • HB1300 enacted law and history ↗

    Official 2026 election source index retrieved from government HTML via text proxy; links enacted law, prior versions, majority/minority reports and House debate. Readers can inspect arguments and changes directly.

    Retrieved / checked October 24, 2026.
  60. [60] NHED • Valuations, assessments and rates, 2007–08 ↗

    July 1, 2008. Original eight-page agency PDF text retrieved through a proxy. Final statewide row: local education tax assessment $1,377,385,153. Compared with the 2025 table’s $2,790,227,910, nominal growth is 102.574%, rounding to 103%. Repeated cooperative reference rows are excluded from totals.

    Retrieved / checked October 24, 2026.
  61. [61] BLS via Federal Reserve Bank of St. Louis • CPI-U all items, U.S. city average ↗

    Original BLS series CPIAUCNS, monthly, not seasonally adjusted, redistributed by the Federal Reserve. CSV independently retrieved October 24, 2026: July 2007 208.299; July 2025 323.048. Growth 55.0886%; national consumer prices are a general purchasing-power measure, not a school-specific cost index. For the salary sensitivity, July–June means are FY2001 175.091667 and FY2024 309.570083. This fiscal-year adjustment is separate from the July-to-July tax comparison.

    Retrieved / checked October 24, 2026.
  62. [62] NCES • National Public Education Financial Survey, FY2001, version 1b ↗

    Primary tab-delimited state file independently downloaded October 24, 2026. NH row: R3 state $884,875,461; STR1 local $751,906,966; STR4 federal $77,364,803; TE11 expenditure field $1,690,009,689; MEMBER00 208,461. Whole dollars. Original version 1a was also checked and these values agree. TE11 alone is narrower than the ELSI/Digest total, which also adds E4D and long-term debt interest E7A1. See the ELSI methodology source.

    Retrieved / checked October 24, 2026.
  63. [63] NCES • National Public Education Financial Survey, FY2019, version 2a ↗

    Primary state file read independently. NH: state $1,007,938,849; local $2,101,425,736; federal $165,688,053; TE11 expenditure field $3,262,591,802; finance-aligned membership MEMBR18 174,581. Earlier version 1a agrees on the revenue totals. Use the finance-file membership, not the broader 178,515 public-school fall count. TE11 alone is narrower than the ELSI/Digest total, which also adds E4D and long-term debt interest E7A1. See the ELSI methodology source.

    Retrieved / checked October 24, 2026.
  64. [64] NCES • National Public Education Financial Survey, FY2020, version 2a ↗

    Primary state file read independently. NH: state $1,055,087,086; local $2,165,951,582; federal $161,745,104; TE11 expenditure field $3,233,054,684; finance-aligned membership MEMBR19 173,124. Local revenue includes more than property-tax collections. TE11 alone is narrower than the ELSI/Digest total, which also adds E4D and long-term debt interest E7A1. See the ELSI methodology source.

    Retrieved / checked October 24, 2026.
  65. [65] NCES • National Public Education Financial Survey, FY2024, provisional version 1a ↗

    Primary state file read independently October 24, 2026. NH: state $1,144,995,592; local $2,555,320,338; federal $335,792,202; TE11 expenditure field $3,976,038,155; finance-aligned membership MEMBR23 159,721. Provisional release. The documented coverage excludes independent charter districts and one district with missing financial reporting; pupil membership is adjusted accordingly. TE11 alone is narrower than the ELSI/Digest total, which also adds E4D and long-term debt interest E7A1. See the ELSI methodology source.

    Retrieved / checked October 24, 2026.
  66. [66] NCES • FY2024 financial-survey documentation and state coverage notes ↗

    2026 release, publication 2026-009. Primary documentation independently read. Membership is adjusted to match finances; NH excludes independent charter districts and one district lacking FY2024 financial reporting (Appendix H). Therefore the expenditure series is reported-system growth, not an identical-district panel. Avoid adding debt principal to a resource-cost measure without explaining double counting.

    Retrieved / checked October 24, 2026.
  67. [67] BEA via Federal Reserve Bank of St. Louis • PCE price index ↗

    Monthly PCEPI CSV retrieved October 24, 2026. This audit aligns prices to the fiscal/school year by averaging July through June. FY2001 74.719250; FY2019 102.793417; FY2020 104.054750; FY2024 122.162083. Real change = nominal ratio divided by the index ratio, minus one. These consumer prices are not a school-input cost index. Calendar-year averaging produces different results; revisions are possible.

    Retrieved / checked October 24, 2026.
  68. [68] NCES • Digest 2002, Table 78 • average teacher salary ↗

    Prepared July 2002; primary federal publication of NEA salary data. NH current-dollar mean salary in school year 2000–01: $38,301. Read the 2000–01 column rather than the adjacent 1999–2000 figure. This is average gross salary, not total employer compensation or the pay trajectory of a fixed group of teachers.

    Retrieved / checked October 24, 2026.
  69. [69] NEA Research • Rankings 2024 and Estimates 2025, Table B-6 ↗

    April 2025; Table B-6 reports NH 2023–24 mean classroom-teacher salary $67,170, without an NEA-estimate marker. Retrieved in full through a text proxy. NEA is an advocacy organization and also the original survey producer whose salary series NCES publishes; its policy views are not treated as evidence. An independently read preliminary NHED district report gives $67,096.40, a different reporting version.

    Retrieved / checked October 24, 2026.
  70. [70] NHED • Teacher average salary, school year 2023–24 ↗

    Report 20, preliminary, November 29, 2023. Primary PDF read through text proxy. State mean $67,096.40 over 13,089.10 teacher FTE. Preliminary district reporting differs from the later NEA series; do not silently substitute one endpoint into the other.

    Retrieved / checked October 24, 2026.
  71. [71] NHED • Public assessment data, 2024–25 ↗

    Primary CSV: 90,395 records; timestamp August 19, 2025. Current and archived copies matched. Selected District Level / All grades / All students / Regular denominator / math and reading. Three suppressed districts and one with only >90% outcomes are excluded from 162 numeric-cost matches, leaving 158. Published percentages are rounded and tested counts are in bands.

    Retrieved / checked October 24, 2026.
  72. [72] NHED • Assessment data definitions and download directory ↗

    Primary definitions inspected October 24, 2026. Regular denominator uses tested students. Proficiency is performance at levels 3 or 4. Rates, raw scores, grade-level status, growth and the percentage of unique students proficient in both subjects are different measures.

    Retrieved / checked October 24, 2026.
  73. [73] NHED • K–12 lunch eligibility, October 2024 ↗

    Final report dated December 2, 2024. Primary HTML retrieved through text proxy. School-level counts reconciled to district totals. All 39 districts in each original cost group report numeric eligibility. Simple mean of district percentages, not pupil-weighted; eligibility is an administrative measure, not a complete poverty census.

    Retrieved / checked October 24, 2026.
  74. [74] NHED • K–12 lunch eligibility, October 2025 ↗

    Final report dated February 3, 2026. Primary HTML retrieved through text proxy. School-level counts reconciled to district totals. All 39 districts in each original cost group report numeric eligibility. Simple mean of district percentages, not pupil-weighted; eligibility is an administrative measure, not a complete poverty census.

    Retrieved / checked October 24, 2026.
  75. [75] NHED • Original FY2024 district revenue/expenditure release ↗

    January 7, 2025 release, independently recovered. Coverage explicitly excludes one nonreporting district. Net state sources $1,150,714,783; local tax $2,446,703,088; tuition/food $67,622,977; other $12,567,723; federal $335,792,202. Total expenditures $3,980,451,955 plus separately listed debt principal $79,041,383. This is the original release behind the 2025 policy report, not the later revised file.

    Retrieved / checked October 24, 2026.
  76. [76] BEA / Federal Reserve ALFRED • PCEPI as available April 1, 2025 ↗

    Original monthly BEA price series preserved by the Federal Reserve. January 2001 74.932; January 2020 104.419; January 2024 122.115. The 2024/2001 ratio reproduces the policy report’s inflation-restated 2001 chart baseline to the nearest dollar. The report names PCE but does not specify months; this exact numerical match reconstructs its timing convention. Later revisions and July–June averaging produce different results.

    Retrieved / checked October 24, 2026.
  77. [77] NEA Research • Rankings 2023 and Estimates 2024, original 2024 edition ↗

    September 2024 report, retrieved in full. Estimated 2023–24 average classroom-teacher salary for New Hampshire: $67,170, also reported in the later 2025 edition. This confirms that a salary revision is not needed to reproduce the original +8% calculation. NEA is the original survey producer, as well as an advocacy organization.

    Retrieved / checked October 24, 2026.
  78. [78] NHED • iExplore, 2025 district economic-disadvantage indicator ↗

    Original live agency dashboard retrieved October 24, 2026. The public wrapper embeds dashboards.nh.gov/t/sonhpublic/views/iExplore/Explore-Dashboard; the older singular dashboard.nh.gov URL has moved. Exported the visible summary for Indicator Grain = District, Year ID = 2025, Indicator Name Clean = Economically Disadvantaged Students. Read ATTR(DownloadValue), retaining suppression labels. In the original lowest/highest-cost 39, 29/27 numeric values sum to 753.71/878.36 percentage points; their means are 25.99%/32.53185185%. No suppressed values were inferred or unmasked.

    Retrieved / checked October 24, 2026.
  79. [79] NH School Facts • original statewide page ↗

    Main page captured October 24, 2026. Original claims, links, wording and caveats inspected directly. This is primary evidence of what the publisher says, not independent confirmation of its factual or causal claims. The assessment excludes authors’ motives and does not attribute ballot-language claims to this site.

    Retrieved / checked October 24, 2026.
  80. [80] NH School Facts • original share graphics ↗

    Share page captured October 24, 2026. Used to identify individual numerical anchors and framing. The score evaluates statewide content, not videos or every municipal output. Original numerical claims receive credit wherever the primary-source reconstruction succeeds.

    Retrieved / checked October 24, 2026.
  81. [81] BEA / Federal Reserve ALFRED • PCEPI as available March 1, 2023 ↗

    Monthly series retrieved independently. January 2001 79.265; January 2019 108.977. Their ratio 1.3748438781303223 exactly matches the inflation multiplier recovered from the publisher’s district workbook; +37.4844% rounds to 37.5%. Historical data vintages matter; the index has since been rebased and revised.

    Retrieved / checked October 24, 2026.
  82. [82] NCES • Elementary/Secondary Information System, fiscal definitions ↗

    Live public table-generator metadata inspected October 24, 2026. Total expenditures (Digest calculation): TE11 + E4D + E7A1. E7A1 is long-term debt-service interest, excluded from TE11; E4D is direct state support for private-school students and is zero for NH in the inspected years. Thus FY2001 total is $1,690,009,689 + $33,014,855 = $1,723,024,544; FY2019 $3,262,591,802 + $44,059,041 = $3,306,650,843. The earlier audit’s TE11-only benchmark was narrower than the published total and has been corrected.

    Retrieved / checked October 24, 2026.
  83. [83] NCES • School District Finance Survey, FY2001 v1d and FY2019 v2a ↗

    FY2001 version 1d (March 2005), paired with FY2019 version 2a at https://nces.ed.gov/ccd/Data/zip/sdf19_2a.zip. Join NH records by LEAID; retain nonmissing state/local/federal revenue in both years. Exactly 170 records: 158 with positive V33 membership in both years, plus 12 without positive membership. Whole-dollar sums from the source fields: state $877,999,000→$1,000,427,000; local $792,004,000→$2,191,912,000; federal $75,783,000→$164,228,000. Matched positive membership totals 203,448→169,074. All reported older funding endpoints reproduce with the original inflation and statewide denominators. A 32-bit float conversion reproduces the printed aggregate precision; the software that performed that conversion is not established.

    Retrieved / checked October 24, 2026.
  84. [84] NCES • FY2019 School District Finance Survey, version 2a ↗

    2022 revised release. Primary file independently matched to FY2001 v1d by stable district identifier. This is the source of the 170-district revenue reconstruction, not a substitution of newer state-finance totals or Census summary tables. See the paired F33 source for criteria, units and totals.

    Retrieved / checked October 24, 2026.
  85. [85] Scafidi / Josiah Bartlett Center • published district appendix workbook ↗

    Original Numbers workbook linked from the publisher’s March 2023 report page, recovered and read. Its 158 district records have positive membership at both endpoints; all FY2019 district expenditure rows match F33. The inflation multiplier recovered from every row is 1.3748438781303223, exactly January 2019/January 2001 PCE in the March 2023 data vintage. The appendix panel and statewide headline totals have different scopes. This workbook establishes the author’s calculations; primary F33 independently confirms the inputs.

    Retrieved / checked October 24, 2026.
  86. [86] NHED • District fall enrollment, October 1, 2019 ↗

    Primary report dated February 20, 2020, retrieved through the agency PrintReport endpoint. Traditional district total 167,584, exactly the FY2020 policy report denominator. SchoolYear 2020 means school year 2019–20, not fall 2020.

    Retrieved / checked October 24, 2026.