What you can ask this archive

Generated by scripts/build_question_bank.py. Do not edit.

108 questions this archive can answer, in plain English. Every one has been run against the data, so nothing on this list is a question we merely hope is answerable.

Some carry a second line in italics. That line says what the answer does not tell you, and it is usually the more important half: a figure quoted without it is the kind that cannot be defended when somebody asks where it came from.

The same list with the query behind each question is at questions.md, and anyone can run those against the data at https://lunenburgbudgetproject.org/api/query.

The school budget

• What did the district budget in total, in each year and at each stage? A STAGE is not a period. proposed, settled and actual are three different documents about the same year, and they must never be mixed.
• Which budget lines grew fastest across the years the archive holds? Both ends are the SAME stage. A rate measured from an actual to a budget is partly growth and partly the step between them.
• Which lines do two documents state differently, and by how much? The documents disagree with themselves by up to 1.5%, which is larger than most variances anybody wants to measure.
• What does each FY27 scenario total?
• How many budget lines are there in each section of the budget?
• Which function groups hold the most budget lines?
• What is in the FY27 workbook, by column? Filter row_kind='line': the sheet's own TOTAL rows are loaded too, and summing without the filter double-counts roughly fourfold.
• Which budget lines appear in the workbook but not in the line catalogue?
• How many figures rest on a document that two sources report differently?
• What is the total salary line in each year, and does the stage change it?
• And total expenses?
• What is the biggest single line in the budget, in each year?
• How many lines does each budget document state?
• Which lines exist in one scenario but not another?
• What does the FY27 workbook say each line was in FY25?

Special education

• How many special education paraprofessionals were budgeted, by school and year? This is the line the 12.8% escalator rests on, and it is dollars, not people.
• And special education teachers?
• What has out-of-district tuition done, year by year?
• How many children were placed outside the district, and where? A count of children placed. It says nothing about which fund paid or what a placement cost, so it does not settle the money.
• Do the placement counts tie to their own parts, and to the prior year?
• What has special education transportation cost, by year?
• Does each year of the placement series agree with what the next report says of it? Two checks travel with this series: the parts sum to the total, and each year states the previous year's figure. n/a is the first year, which has nothing before it.

The town's books

• What did each department spend against its budget, in the latest period held?
• Which departments are spending faster than the year is elapsing?
• What funds does the town keep, and what restricts them?
• What moved through the special revenue funds in each year?
• How many accounts are there at each level of the chart?
• Which grants did the district receive, and who owns them?
• What does the ledger hold for each fiscal year and period? Period 13 is the year-end close, after purchase orders are cleared. Period 12 is not the end of the year.
• Which function codes can be compared between the budget and the ledger? This is the level at which the two systems join. Below it they do not: the town shortens account names to ten characters.
• How much did each function group budget and spend across all years?
• What went through fund 1301, and when?
• Which accounts had the largest unspent balance at the latest period?
• How much was transferred into or out of each account? transfers is CUMULATIVE. Movement between two periods is the difference of the column, never the later value.
• Which accounts are revenue rather than expenditure? Revenue rows are stored NEGATIVE, exactly as MUNIS prints them. Check account_type before doing arithmetic across types.
• What share of its budget had each department used?

Staff on the rosters

• How many people of each kind did the town print on a roster, by year? A count of names the town printed. It carries no FTE and no funding source, so it is not a staffing level.
• How many paraprofessionals, by school and year?
• How many paraprofessionals were tied to a named grade?
• What did the town print as the title for a paraprofessional, year by year? Tutor, Aide, Tutors/Aides, Paraprofessional, Para, (para), Sped Para. Five names for one job across fifteen years, which is why role_category exists.
• Which roster titles could not be classified at all? Left unknown rather than guessed. 8% of rows.
• Which rule decided each classification, and how much rests on the weakest ones? A rule beginning heading- read the section heading rather than a printed title, which is weaker evidence.
• How many names appear on each school roster in each year?
• How many teachers were printed against a specific grade?
• Which schools have roster entries, and for which years?
• How many counselors, nurses, psychologists and social workers, by year?
• How many people did the town print on a roster in total, by year?
• Which names appear across the most years? A name printed in a public annual report. It is not a claim about employment, and a roster gives no FTE.
• How many administrators did each school print?
• Which grades appear anywhere on the rosters?

Athletics and fees

• What did each sport cost, and how many played?
• Do the per-sport figures add up to the totals the district printed?
• What has athletics cost and raised, year by year?
• What has the athletic fee been, by year and tier?
• Which rates does this project know about, and which does it use? It deliberately includes rates the model does NOT use, and the ones that cannot be stated at all.
• Which rates were set by a document we hold, and which were not?

The annual town reports

• What appropriations did each annual report print, and are the rows checked? ALWAYS split on status. checked, check failed and no check are three different claims and nothing may be aggregated across them.
• What did the town pay in gross wages, and to how many people?
• What debt has the town carried?
• What capital projects did the reports list?
• What is in the trust funds?
• What did the town value its property at?
• What enrollment and MCAS results were printed?
• What did the town assess for Monty Tech?
• What is in the Monty Tech pages of the annual town report?
• How many births, deaths and marriages were recorded?
• Who held town office, and when?
• What did each department report doing?
• What receipts did the town record, and from what source?
• What tables does each annual report contain?
• Which tables in the reports were read, and which are still uncaptured?
• What did the survey find on each page of each report?
• What is catalogued in each report, by printed heading?
• Where did the extraction find something it could not reconcile? An anomaly is a finding about our reading of the page as much as about the page.
• Which special revenue funds appear in the reports, and in which years?
• How many figures did each annual report yield, and how many were checked?
• Which report tables print a total we can reconcile to?
• How many pages of each report carried figures at all?
• What kinds of anomaly did the extraction find, and how often?

Revenue, tax base and free cash

• What free cash has each town certified, and from what? Absolute dollars with no denominator, so they do not compare between towns of different size. The composition does compare, because a share has no size.
• How does Lunenburg free cash compare with its neighbours, by composition?
• What has the town spent on capital, and where did the money come from?
• What is in the FY27 capital plan, and what is funded?
• How do state measures compare Lunenburg with other districts?
• Which DESE measures reconcile against the printed totals, and which do not?
• How has free cash moved for Lunenburg specifically?
• Which towns does the free cash comparison cover?
• What measures does the state publish about this district?

Votes and elections

• What has the town been asked to fund, and did it agree?
• What turnout did each ballot question draw?
• What election results did the annual reports print?

Provenance, and what is not established

• Where did a figure in this dataset come from? This is the join that gives an annual-report row an address. Use it in any query whose answer somebody might want to check.
• Which documents does the archive hold, and how were they obtained?
• Which documents no longer open at the publisher, or no longer match our copy?
• Which documents have no upstream address at all? A gap on our side, not the town's: they were gathered before the mirror existed and nobody wrote down where they came from.
• Where do two sources state the same budget line differently?
• How much of each dataset has been checked against the page it came from?
• Which figures has somebody stated publicly, and on what basis?
• Which budget lines have no ledger account mapped to them? The crosswalk is empty ON PURPOSE. District lines are named, MUNIS rows are coded, and no published document maps one to the other. Budget-to-actual at line level cannot be answered from this data.
• Which pages of the annual reports have columns we could not establish? v1 is an ordinal -- the first column of THIS page that held figures -- not a column name. Read column_meaning before summing anything.
• How many rows of each report table failed their own check?
• Which documents were obtained by records request rather than published?
• What basis does each document have for the figures it prints? ledger means a figure exists because a transaction did. restatement means a prior year re-presented by the party that spent it. They are not interchangeable.
• Which datasets have a document for every edition, and which do not?
• How many rows of each dataset came off each page?

Finding your way

• What tables are there, and how big are they?
• What fiscal years does the archive cover, per dataset?
• What does one fiscal period mean?