Municipal spending explorer

About the data

Spending per pupil is the Massachusetts Department of Elementary and Secondary Education's (DESE) per-pupil expenditure (CSV) from each district's end-of-year report, divided by in-district FTE and grouped into the report's nine categories; its Other combines DESE's professional development and instructional materials, equipment, and technology. FY2005 to FY2008 come from DESE's workbooks for those years, which its Public Records Center provided in October 2026 in answer to public records request P004939-101625 (the files, and how they were read). It leaves out tuition for pupils educated elsewhere, because, as DESE puts it, those costs "vary greatly depending on the reason for going out of district." The panel Pupils educated elsewhere shows them on their own, per out-of-district FTE pupil. DESE's data has no separate figure for them from FY2009, so the explorer works it out from DESE's totals: total spending less in-district spending. For FY2005 to FY2008, where DESE printed the figure, the two agree to within 0.7%.

Enrollment comes in three measures. In-district FTE is full-year enrollment of students who live in the district or are choiced or tuitioned into it. Foundation enrollment counts the residents the district is financially responsible for on October 1 of the prior year. Out-of-district FTE counts the district's pupils educated elsewhere. In-district and out-of-district FTE come from DESE's expenditure data (CSV), foundation enrollment from its Chapter 70 district profiles (CSV).

Chapter 70 aid and net school spending are per foundation pupil, from DESE's Chapter 70 district profiles (CSV). The state sets a required level of net school spending for each district each year. Chapter 70 aid pays part of it, and the town, or a regional district's member towns, pays the required local contribution. Anything above that is a local choice. Because spending per pupil and net school spending use different denominators, the two charts are not the same measure: they are a picture of apples in the morning compared with a picture of apples in the evening. Both dollar charts share one scale.

Since 1990 adds FY1990 to FY2004, from DESE's per-pupil reports in the same public records request. For those years DESE measured only what it called day-program spending, an older, narrower measure: what a district spent running its own schools, mainly from its school committee budget and the town's indirect costs for them, divided by the pupils enrolled there. It leaves out pupils sent to other schools and what is paid for them, so it usually runs below the columns. It continues to FY2006, so the two measures overlap for two years: in FY2005, which DESE's workbook sets out both ways, Abington's was $6,766 per pupil against $7,463 in the columns. The explorer draws it as a separate dashed line, and each measure has its own trend and growth rate. The older measure's averages weight each district by DESE's average membership from FY1997. DESE's reports give no pupils for FY1990 to FY1996, so those years use the fall enrollment the National Center for Education Statistics counts (Common Core of Data), a head count from a different agency; in FY1997 and FY1998, when both exist, nine districts in ten differ by less than 9%. Chapter 70 figures reach back to FY1993, DESE's first profile year. Districts that no longer exist, including towns that once ran their own schools, are listed under Former districts with the years they have figures for.

Totals, under Amounts, turn the figures per pupil or per resident back into dollars. A district's spending and revenue per pupil are multiplied by its in-district FTE, and its Chapter 70 figures by its foundation enrollment. A town's figures are multiplied by its population. An average's totals are its average member's: the group's pooled figure times its average enrollment or population. A total moves with enrollment and population as well as with spending, so it can rise while spending per pupil falls.

Cohorts group traditional districts by FY2025 in-district FTE, using the boundaries the report set from FY2024 Western MA enrollment: Tiny up to 200 FTE, Small 201–800, Medium 801–1,600, Large 1,601–10,000, and Very large above that. Vocational districts and charter schools are groups of their own. Regions are the state's six EOHHS regions. As the report left Springfield out of its Western MA average, region averages leave out the ten very large districts, which have an average of their own; together they teach nearly a quarter of the pupils in traditional districts. There are two statewide averages, without them and with them, and the same for towns, without and with the ten cities they serve (groups and their members). A region with fewer than three districts in a cohort has no regional average for it, so "its cohort in its region" uses the statewide cohort there. Spending averages are weighted by in-district FTE, Chapter 70 averages by foundation enrollment, and enrollment averages are per district. Gateway Cities are the 26 mid-size cities the Commonwealth lists, whose median household income and college attainment fall below the state's (the list dates from 2014). Their districts, and the cities on the Towns tab, have averages of their own, alongside their size cohorts and peer groups. Two virtual schools, the state's institutional schools, and the Massachusetts Academy of Math and Science are not included.

Community types follow the Commonwealth's own lines where it draws them. A traditional district is rural at 35 or fewer students per square mile, its foundation enrollment over its towns' land area: the test for DESE's rural school aid, which also requires income below the state average. DESE's three tiers, 11 or fewer, over 11 to 21, and over 21 to 35, have averages of their own. A town is rural under 500 residents per square mile, the line in the Rural Policy Plan for the Commonwealth's Rural Policy Advisory Commission. The Inner Core is Boston and its densest neighbors, the Metropolitan Area Planning Council's Metropolitan Core Communities and Streetcar Suburbs in its community types, as its DataCommon profiles give them: Boston, Cambridge, Chelsea, Everett, Malden, Revere and Somerville, and Arlington, Belmont, Brookline, Medford, Melrose, Newton, Waltham, Watertown and Winthrop. The other 22 Gateway Cities are a type of their own; Chelsea, Everett, Malden and Revere count with the Inner Core, though the Gateway Cities comparison keeps all 26. The state has no name for the rest, called suburban and other here.

Income fifths rank the 351 towns by income per resident: the residents' income the Chapter 70 formula uses for FY2027, from 2023 state tax returns, divided by the Census Bureau's population estimate for July 2023. Each fifth holds 70 or 71 towns. A district's fifth ranks the combined income per resident of its towns among the 286 traditional districts, 57 or 58 to a fifth. Each fifth has an average of its own, for "Its income fifth" under Compare with. The cap tab ranks the towns the same way, for whichever year it shows. Land areas are the Census Bureau's (2023 cartographic boundaries); enrollment is DESE's latest foundation enrollment, and population the latest estimate. A list of every group and its members has a page of its own, and a file (CSV).

Revenue by source comes from the Census Bureau's Annual Survey of School System Finances (CSV, in dollars, each column naming the Census items it adds up), divided by in-district FTE like spending. It covers regional districts only: for many town-run districts, Census's local revenue is only part of what the town spends on its schools (for Amherst in FY2024, $18.4 million of revenue against $37.6 million of spending). Payments from towns include a regional district's member-town assessments; payments from other school districts include tuition for pupils they send. Revenue per pupil usually runs about a third above spending per pupil, because it also pays for what the spending measure leaves out, such as tuition for pupils educated elsewhere, school buildings, and the teachers' pensions the state pays into the teachers' retirement system on districts' behalf. School building aid can arrive in one large payment. In FY2009 the state paid $412 million of Chapter 70 aid with federal stimulus money, so Census counts about a tenth of that year's Chapter 70 aid as federal. Census publishes each year's figures more than a year after it ends.

Town budgets come from each town's Schedule A (CSVs: spending, revenue, each with population), the year-end report of general fund revenue and spending that town accountants file with the state Division of Local Services. Enterprise, capital, and other funds are not included. Each town's school districts and FY2025 required local contributions come from DESE's Chapter 70 tables (CSV), and each town takes the region of its own school district, or of its regional district. Peer groups put each town with others its size, by the July 2024 population estimate: under 1,000 residents (26 towns), 1,000–2,499 (42), 2,500–9,999 (98), 10,000–24,999 (103), 25,000–49,999 (55), and 50,000 or more (17). The ten cities whose school districts are very large, from Boston to Springfield, stand apart from the town averages, as their districts do from the district averages. Averages pool their towns per resident. Dollars are per resident, using the Census Bureau's population estimate for July 1 at the start of each fiscal year. They can also be shown per household, using the Census Bureau's American Community Survey five-year estimates (CSV), or per taxable property, using the parcels of taxable real estate on each town's LA-4, or in total. For FY2021 the April 1, 2020 estimate stands in, because the July 2020 estimate drops by more than a quarter in Amherst and recovers a year later.

Property comes from the state Division of Local Services. Parcels and assessed values are each board of assessors' yearly LA-4 report (CSV), as of January 1 before each fiscal year, grouped by the state's property use codes. Multifamily is two- and three-family homes and apartment buildings (codes 104, 105, 111–125). Other residential is mobile homes, lots with more than one house, and vacant or accessory residential land (103, 109, 130–132, 106). Mixed use is property with more than one use, such as a store with apartments above it (012–043). Farmland is land taxed under the Chapter 61A agricultural program; forest, recreation, and open space is land under Chapters 61 and 61B and class 2 open space. Personal property, such as business equipment and utility lines, is assessed but has no parcel. Exempt property, such as schools, churches, and public land, is left out. The average single-family tax bill is the state's: single-family assessed value divided by single-family parcels, times the residential tax rate, set beside the Department of Revenue's income per capita. Amherst's FY2025 bill is 41% of its income per capita, the highest share in the state; income per capita counts every resident, students included.

The property tax levy is what a town raises from property taxes, and Proposition 2½ (CSV) limits it. Each year a town's levy limit grows by 2.5% plus the tax on new construction, and voters can raise it with an override or exclude a debt from it; the charts' limit includes those exclusions. The limit itself can't pass the levy ceiling, 2.5% of the town's assessed value, though excluded debt can. Unused levy capacity is the gap between the limit and the levy a town sets: money it could raise without asking voters, and chooses not to. It can be read as tax relief, as careful budgeting, or as austerity. The tax rate here is the single-family bill over the home's value, which is the residential rate except where a town gives a residential exemption.

Equalized valuation (EQV) is the state's estimate of the full and fair cash value of all taxable property in each city and town on January 1, set every two years by the Commissioner of Revenue (EQVs since 1970; CSV). Assessors value property at different distances from market prices, so the state's Bureau of Local Assessment compares each town's assessments with its recent sales, class by class, and divides each class's assessed value by that ratio: homes assessed at 95% of their sale prices have their value divided by 0.95. It then adds an estimate of new construction (DLS; “EQV: Leveling the Playing Field”). Chapter 70 uses EQV to measure each town's property wealth when it sets what the town should contribute. In FY2025, a town's target local contribution was 0.3906% of its January 2022 EQV plus 1.4312% of its residents' 2021 income. The state chose the two percentages so that, statewide, property and income would each supply half of a total equal to 59% of all foundation budgets. No town's target could exceed 82.5% of its foundation budget, and 182 towns were at that cap. Each year a town's required contribution moves from the year before toward its target, and Chapter 70 aid pays at least the rest of the foundation budget.

Town and school budgets overlap without lining up. Education includes the town's own school department and what it pays regional and vocational school districts, so it does not match any one district's spending, and health insurance and pensions for a town's own school employees usually sit in fixed costs. State aid includes Chapter 70 aid for a town's own school district; aid to a regional district goes to the district. Intergovernmental assessments are the state and county charges on each town's Cherry Sheet, such as charter school and school choice tuition. Required local contributions are what the Chapter 70 formula sets for each town and district in FY2025.

Revenue groups combine Schedule A's eleven sources into six: charges, fees, and fines are service charges, licenses and permits, fines and forfeitures, and special assessments; other revenue is miscellaneous revenue and revenue from other governments; transfers and other financing are transfers and other financing sources. Figures are as each town filed them, and a few look like one-time entries: Lee's FY2024 revenue includes $27.3 million of miscellaneous revenue, more than half its total.

FY2025 dollars adjust for inflation with the Bureau of Labor Statistics' consumer price index for the Northeast (CSV; CPI-U, series CUUR0100SA0), averaged over each July–June fiscal year. The report's inflation addendum restates its growth rates against this index and against the federal price index for state and local government purchases.

Source files

The agencies' own files behind this explorer are on GitHub, in the repository with the code that prepares them. Each file below links to its copy there, and each source links to the agency's page for it. The files keep the rows as each agency published them: the scripts that download them keep the Massachusetts rows of national files, stack years into one table, and add DESE's district codes. The repository's data README names the script that downloads each file again.

The CSV links in the paragraphs above are this explorer's own files: the data for these districts and towns as each agency published it, except that the revenue file adds Census items up into the chart's sources, and every row links to its source. Built by Tim Shores with Claude Code. The code that prepares the data is at github.com/timshores/schools.