August 30, 2026

Spending free cash and passing an override are not the same thing. The same $794,872 defers one year as free cash and is worth $4,396,563 more over six as an override — and neither closes the gap, because both grow slower than the cost of running the schools.

• —The free cash page can now apply the money to the projected gap, and it is OFF until you switch it on. Nothing anywhere else on this site changes: the projection is built without free cash, and the figure it produces is unchanged whether the switch is on or off.
• —Side by side with an override of the same size. Free cash closes FY28 and nothing after. An override lifts the levy limit permanently and the schools keep it every year, growing at the 2.5% cap. Over six years the override is worth $4,396,563 more from identical dollars.
• —And the row that matters is the last one, not the total. Even the override leaves $3,740,932 of gap in FY33 and the shortfall grows every year, because an override rises at 2.5% and the cost of the schools rises faster. Permanent money loses ground more slowly than one-time money, and still loses ground.

August 30, 2026

A new page lets you draw the town’s free cash down to any level and see what it covers. Emptying the reserve entirely — which nobody proposes — defers the school budget gap two years and leaves the town with nothing.

• —The state’s certified free cash for Lunenburg and eight comparable towns, 2021 to 2025, is published with the line-by-line calculation behind it. Lunenburg certified a record $3,354,370 for this year, 6.55% of the operating budget on our figures and 6.65% on the Town’s.
• —Two claims are argued about locally — that the town is too conservative, and that free cash is not up to standard so the town is rebuilding. The Town’s own budget release contains both and both are true: a record year, and below the recommendation in seven of the last ten. They describe different windows.
• —What the data adds is that a normal year does not refill it. The record exists because unspent appropriations were $2,457,761 against a four-year average of $986,340 — 2.49 times, the largest jump of nine towns, while two of them fell. Put that one line back at its own average and the town certifies 3.96%, below the bottom of the band it is measured against.
• —And free cash cannot bend the curve. It is one-time money: a level, not a rate. Drawing down to the 5% floor releases $794,872, which covers FY28 and nothing after it. The page shows the whole ladder, including below the floor, because "what if we spent it all" is the question people actually ask.

August 30, 2026

Two claims are being made about the same number — that the town is too conservative with free cash, and that its free cash is not up to standard and is being rebuilt. The state’s own calculation for Lunenburg and eight comparable towns is now published. It does not settle the argument, and the analysis says why in its first paragraph.

• —The Division of Local Services publishes a free cash proof for every community — the year-end calculation of what a town may appropriate without raising taxes. Lunenburg’s and eight neighbours’, 2021 to 2025, are now in the archive with the line-by-line detail. They reconcile to themselves twice over, in 81 checks that all tie to the dollar.
• —The disagreement cannot be settled from them, because a standard for free cash is a ratio and these files are a numerator. There is no population, budget or revenue figure anywhere in them. Lunenburg’s balance can be set against Lunenburg’s own budget — 6.51% of a $51,531,199 general fund — but not against any neighbour’s, and the standard being invoked is in a document this project has not read: the town’s own Financial Policies Manual.
• —What the figures do show is the shape, and the shape has moved. Two thirds of Lunenburg’s 2025 free cash is money that was appropriated and never spent — $2,457,761, against $1,225,720 from revenue coming in above estimate. That share was 31% in 2023. A balance built from underspending is a different thing from one built from revenue beating forecast, and the two imply different remedies.
• —None of that means waste, and none of it is the schools: every town turns money back, and this is a town-wide figure across all 67 departments with no breakdown published. Which departments turned back the money, and whether it is the same ones each year, is the difference between a pattern and a run of one-offs.

August 30, 2026

Two years of every town board’s agendas and minutes — 1,383 documents across 40 boards — are now published as full text rather than as an index of dates. Zoning, conservation, health, planning, cemeteries, the library, housing, the schools. It is where the town argues, and none of it appears in a budget document.

• —The text of every agenda and set of minutes is readable and downloadable. Until this build the site published only an index pointing at the town’s scanned PDFs, which meant that somebody searching this site for what the School Committee said about a contract, or what the Zoning Board decided about a property, found nothing and reasonably concluded it was not here.
• —Each board is also published as a single file — the whole School Committee in one 0.9MB document — because you cannot search a website but you can read one file. Every document inside carries its own permanent address, so anything found there can be cited to the document rather than to the bundle.
• —The site now states, on every page, what it holds and which file answers which question. That is for anybody pointing an AI assistant at it: the assistant gets the inventory, the download addresses, and the one warning that matters — that this archive holds both what the town budgeted and what it spent, and mixing them produces a confident wrong answer.
• —It also says plainly that this is not only a budget site. An assistant told otherwise will answer "this cannot help you" to a question about zoning or conservation that the archive answers completely.

August 30, 2026

Athletic fees for FY26 were modelled at $250 a season when the district had voted $325. Not a wrong number — a right number from the wrong year, taken from a fee schedule that states its rates and never states which year they cover. Every fee figure on the site has moved.

• —A new page, Rates, fees and contracts, listing every rate this analysis knows about with the fiscal year it applies to, the document that set it and the date it was set. 62 rates; 37 of them checked against a spreadsheet cell or a direct quotation. It deliberately includes the ten rates we cannot state at all, because a fee the town charges and does not publish is a finding rather than a blank.
• —The correction came from a School Committee vote nobody had connected to the model. On 26 February 2025 the committee approved athletic fees of $325 for high school and $275 for middle school, a 25% sibling discount, reduced fees of $50 and $40, and a $1,500 family cap. A fee voted in February 2025 applies to the following school year, which is FY26. The schedule the model was using was correct — for FY24 and FY25.
• —Two figures that had been recorded as unexplained turn out to be the same error. $314.28 per high school participation is impossible under a $250 fee and unremarkable under $325; $233.60 per middle school participation is impossible under $200 and fine under $275. The model needed a 1.452x calibration constant to reconcile its own arithmetic with what the fund actually collected. That constant is now 1.132x, and the remaining gap is stated rather than absorbed.
• —The note on athletic transportation said the line was "budgeted well above what athletics has ever actually spent". That was wrong, and it is the error this project is most careful about elsewhere: those figures are the town’s net appropriation, not what athletics spent. The district’s own workbook puts FY24 athletic transportation at $117,555.00 against a $40,000 line — budgeted BELOW the cost, with the revolving fund carrying the difference.
• —Bus fees are published here for the first time, from the Superintendent’s own emails: $180 for one student, $270 for a family, with reduced and waived tiers. These are fees for getting to school, not to games. They matter to the budget because a bus fee nets down the general education transportation line — so that line can fall without any cost falling.

August 29, 2026

Three years of the athletics revolving fund’s cashbook arrived from the Town — the first complete fiscal years of ledger data this project has held for anything touching school money. Every receipt and payment, with a date.

• —A new analysis, The athletics ledger, reading the fund’s cash for FY2024, FY2025 and FY2026. The years chain end to end against the opening balances the town itself prints. In FY2025, four general journal entries described only as an adjustment "per memo" account for $254,121.18 — 65% of everything that came into the fund that year. What they were for is not established, and the analysis says so rather than guessing.
• —Figures that had been carried as unproven are now sourced. Athletic transportation cost $117,555.00 in FY24 and $91,066.06 in FY25, and both reproduce to the cent from a workbook the Town supplied. What changed is the provenance, not the arithmetic.
• —A missing document now returns a real error instead of quietly showing the home page. Until this build, asking for a source document that was not there answered as though it were — so nothing, and nobody, could tell a document that exists from one that does not. For an archive whose promise is "here is our copy, check it yourself", that was the worst available failure.
• —Every page now sends its content as HTML rather than building it in the browser. A reader with JavaScript sees no difference. A search engine, a link preview, or somebody pointing an assistant at a page to check a figure now gets the page instead of an empty shell.

August 28, 2026

About a fifth of the school budget was folded in with salaries and escalated at the teachers’ contract rate. It now has its own section, and a rate measured across eight to eleven of the district’s own budgets rather than taken from a contract.

• —A section on Bend the Curve, and two findings on the front page. The district’s published cost increase for next year is 3.98%. Inside it, out-of-district tuition falls 46% in a single year — a one-time drop, not a slower rate of growth. Hold that line where FY26 had it and the same budget rises 6.23%. That is the rate to plan against.
• —A contract sets what one person is paid. It says nothing about how many people are employed — and on this line, that is where the movement is. Special education paras are on an agreement giving 2.0% a year and their budget has grown 12.78% for 10 budgets. Teachers are on one giving 3.5% and theirs has grown 2.67%. Both bargained, both wrong, in opposite directions.
• —Out-of-district tuition is no longer escalated at all. Eleven budgets run from $489,918 to $1,291,293 with no direction — the compound rate swings from -46% to +12% depending only on which year you start counting. There is no rate to measure, so the line is held flat and what it would cost to be wrong is published as a table of priced scenarios.
• —Every budget line counted as special education is now listed, with the reason each was counted, because the state has no account code for it and the total is therefore ours. Publishing that list found English Language Learner costs inside it, which are now removed.
• —The rate this line carries has changed four times in a day, and every version is on the page with the reason it was wrong. A number whose history you cannot see is one you have to take on trust.
• —And the decision to hold out-of-district tuition flat now shows its working. It was reached three separate times from sources with nothing in common — eleven budgets, five years of actual spending, and the two halves of the line measured apart — and all three say the same thing: no rate describes it.

August 28, 2026

Every headline figure now names the document it came from, and that document is here to download. If you checked a number before and could not see where it came from, you can now.

• —A sources page listing every document this analysis reads. Each one is downloadable, and the archive is split three ways: published by the town, the district and the state; held for reference but feeding no figure; and written by this project. The third section is the one to read skeptically.
• —Citation markers beside the figures themselves, saying for each whether it was published by somebody, set by contract, fixed by statute, or estimated by us. The estimates are marked as estimates wherever they appear.
• —The district’s budget page and the town’s finance pages mirrored in full, back to FY18, so a document nobody kept is still checkable. Most of it feeds nothing on this site and is labeled that way.
• —Machine-readable copies of the model, the sources index and the budget lines, announced at /llms.txt, so anybody checking this with a tool of their own does not have to scrape the pages.

August 27, 2026

The original build: the FY28 gap, the cut cascade, the levers and the override arithmetic.

• —The five-year projection from the FY27 adopted budget, the priority cascade, the fee and administration levers, and the tax-base pages.
• —Figures were stated without saying which document each came from. That is what v2 fixes.