How we know

Assumptions register

What does the forecast assume, and why?

Every key assumption behind the base case, the evidence for it, what it would be worth by FY2040 if it were wrong, and where to change it in the Forecast Explorer.

Each entry follows the same pattern. Base case is what the forecast assumes. Why is the evidence. If it is wrong is the change in FY2040 room, revenue minus spending, when that one assumption moves and everything else stays at the base case; a minus sign means less room. Change it names the setting in the Forecast Explorer.

Pay

Town staff pay

• Base case: 3.0% a year, the Town's bargaining target.
• Why: It is the figure the Town bargains toward.
• If it is wrong: one point higher for every group of staff, −$42.4M; half a point lower for every group, +$19.4M.
• Change it: "Pay raises compared with the base case".

How much of each budget is pay

• Base case: Pay is 83% of the school budget and 76% of Town department spending. Among Town departments it is 87% of Public Safety, about 85% of Human Services, about 80% of General Government, 63% of Public Works and 52% of Culture and Recreation. Pay grows at each group's rate; the rest of each budget grows with prices.
• Why: Budgeted salaries against actual spending, FY2019 and FY2026; the shares barely moved between the two years.

School pay

• Base case: 3.7% a year, 0.7 points above the Town target.
• Why: The current teachers' contract gives 2.75% raises. The School Department's own budget arithmetic adds step and degree-column increases and subtracts savings from turnover, which comes to about 3.3% to 4.1% a year per position; 3.7% is the middle of that range. The FY2011 to FY2025 record was about 3.4%. The contract expired in August 2026 and is being negotiated.
• If it is wrong: at the School Department's own forecast, about 5% a year, −$32.2M.
• Change it: "School pay above the Town target".

Public Safety pay

• Base case: 3.5% a year, 0.5 points above the Town target, plus the increases already agreed: new fire steps from FY2028, estimated at $909,471 a year, and four police positions refilled with override money, two in FY2028 and two in FY2029.
• Why: Police and fire cost per position grew about 3.3% a year from FY2011 to FY2025, on flat or falling headcount. The faster rise from FY2021 to FY2025 came mostly from one-time changes, a police stipend rising from 4% to 7.5% of pay and new steps, which are already in the FY2027 budget. Contract raises themselves have run 1.5% to 3%.
• If it is wrong: school and Public Safety pay both held to the 3.0% target, +$20.0M.
• Change it: "Public Safety pay above the Town target".

Police and fire staffing

• Base case: the FY2027 budget, 128 sworn police officers and 160 firefighters, with the four police positions to be refilled in FY2028 and FY2029 (see Public Safety pay), plus a fire company of 20 firefighters, at $3.0 million a year in FY2025 dollars, once the town reaches about 31,200 homes. At the recent pace of building that comes after FY2040, so the base case adds none; faster growth reaches it sooner. Police cost grows a little with each new home, through the Town cost of serving a new home (below).
• Why: Fire staffing was 161 from FY2010 to FY2026 and police and fire positions together fell by five from FY2011 to FY2025 while the town's building permits added about 1,250 homes. Fire companies are staffed around the clock, so fire staffing changes in steps. The threshold is the number of homes at which Brookline's firefighters per 1,000 homes would fall to the median of 21 nearby cities and towns (5.16); the cost is the Fire Department's own estimate for a South Brookline ladder company, health insurance and overtime included. When the departments run below their budgeted numbers, the money not spent returns as free cash, which the forecast already counts. Police calls fell about 4% a year from FY2016 to FY2024, and Brookline's officers per 1,000 homes (4.58 in 2019) are close to the same towns' median (4.42), so the same test gives no clear point at which police would step up.
• If it is wrong: police back to 136 and fire back to 168, their levels as recently as FY2018 to FY2020 and in FY2004, −$2.5M; fire cost growing a little with every home instead of in one company step, which the base case doesn't reach before FY2040, −$0.4M; police growing with homes to hold the median, 3 more officers by FY2040, −$0.2M.
• Change it: "Police and fire staffing".

Benefits

Health premiums

• Base case: 12% a year in FY2028 and FY2029, then 8% a year.
• Why: Recent increases have run 10% to 12%; the 25-year average is about 6.5%. The base case settles between the two.
• If it is wrong: 10% a year from FY2030, −$30.4M; 6.5%, +$19.4M; 4.5%, the actuary's long-run trend for retiree health, +$41.4M. These settings move the health insurance line; the retiree health liability follows the actuary's own trend (below).
• Change it: "Health premium growth, FY2030 on" and "Health premium growth, FY2028 and FY2029".

Who is on the health plan

• Base case: Active employees follow staffing. Retirees, about 36% of the plan's cost, grow 0.92% a year on their own.
• Why: Retirees and their survivors are about half of the plan's subscribers (1,659 of 3,390 in FY2025), but many are on cheaper Medicare supplement plans: the FY2024 audit's payments for retirees, $12.4 million, were 36% of that year's group health spending, $34.8 million. Retiree subscribers grew from 1,459 in FY2011 to 1,659 in FY2025.

Pensions and retiree health

• Base case: The pension system reaches full funding in FY2030, on the Town's funding schedule. From FY2031 the Town pays $43.0 million a year into the retiree health (OPEB) trust, as its plan does, on top of retirees' premiums, until the unfunded liability is paid off. Then the money is free again; the base case sets it aside for a decision then and doesn't count it as room. The pension cost that remains after full funding grows 0.65 point a year faster than Town pay, 3.65% in the base case: the actuary's 3.5% payroll growth plus 0.15 point for longer lives. Other benefits follow the payroll of all staff, Town and schools. The liability is rolled forward from the FY2024 audit on the actuary's own trend for retiree health costs, about 5.9% a year in FY2026 falling to 4.5% by FY2035. On that trend it rises to about $224 million in FY2030 and is paid off in FY2036. If retirees' costs grew instead with the base case's health premiums, 8% a year, it would not be paid off before 2040.
• Why: Both are the Town's own adopted schedule and policy, and the retiree health trend is the actuary's. The audited record supports it: from FY2017 to FY2024 the total liability, before subtracting the trust, barely moved, from $292 million to $302 million, and payments to retirees rose about 1% a year.
• If it is wrong: keeping the freed pension money in the budget instead, +$34.6M; counting the money freed by the retiree health payoff as room, +$43.0M. Full funding arriving two years late barely moves the room: FY2030's last catch-up payment is spread over the extra years, paid from the money set aside for retiree health, so that liability is paid down later.
• Change it: "Pension savings after full funding" and "Pension full funding slips".

Schools

School staffing

• Base case: School staff fall half as fast as enrollment, from about 1,242 in FY2027 to about 1,110 by FY2040, while enrollment falls about a fifth. Students from new homes are staffed in full.
• Why: It is what Brookline did: from FY2019 to FY2025 enrollment fell 10.6% and staff fell 5.0%. That is about ten fewer positions a year, which retirements and departures would likely cover.
• If it is wrong: staff held at the FY2027 level, as the School Department's FY2027 to FY2029 plan does, −$27.8M, and the room falls a little short in FY2028 and runs out for good from FY2036; staff follow enrollment down fully, +$9.5M.
• Change it: "School staff as enrollment falls".

Enrollment

• Base case: NESDEC's forecast, scaled to the actual counts for fall 2025 and fall 2026 and updated for actual births, falling to about 5,400 by FY2040.
• Why: The decline follows births. Children entering kindergarten through fall 2029 are already born, and resident births fell from about 590 a year in 2015 to 2019 to 422 in 2023 and 443 in 2024. NESDEC's 2023 forecast had estimated 518 and 507 for those two years, so the base case scales each of those classes, and later ones, down to the births that actually happened, holding births from 2025 at the 2022 to 2024 average, 449 a year. The 2025 count, 423, came in a little below that; the base case keeps the average.
• Alternative: The School Department's own forecast (Cropper-McKibben) has enrollment rising again to about 7,400 by FY2034, mostly because it projects more births (about 600 a year). If births stay near the counted level, that enrollment would take about 1.15 students for every resident birth, against a twenty-year high of 0.90.
• If it is wrong: enrollment rebounds as the School Department forecasts, and staff rise with it, −$47.9M. (Staff fall half as fast as enrollment only while it falls; students added when it rises are staffed in full.)
• More: Enrollment: three forecasts compared.
• Change it: "School enrollment forecast".

Families moving in

• Base case: Families move in and out at the rates in NESDEC's forecast: each class grows about 2% between kindergarten and grade 12, and about 0.85 kindergartners arrive for every resident birth five years earlier.
• Why: NESDEC's rates come from Brookline's own recent history, which they sit between: before COVID classes shrank about 7% from kindergarten to grade 12, and since COVID they have grown about 6%.
• Alternative: Since COVID, classes have grown about 0.35 point a year faster than NESDEC assumes. Brookline's Census blocks show children per home rising about a fifth from 2010 to 2020 within every kind of home, so more families with children can move into the homes Brookline already has. These families add students and cost but no new tax revenue, since their homes are already taxed.
• If it is wrong: each class 0.35 point a year larger, the rate since COVID, enrollment 5,610 in FY2040 (+200), −$4.4M; 1 point a year larger, 6,000 (+580), −$13.0M; 0.5 point a year smaller, 5,150 (−270), +$6.0M. The setting applies to children of every age from birth, so kindergartners per birth move with it.
• More: Aging homeowners and family-sized homes, Students from new homes and Enrollment: three forecasts compared.
• Change it: "Families moving in".

School costs other than staff

• Base case: Grow with prices.
• If it is wrong: two points a year faster than prices, −$8.6M.
• Change it: "School non-personnel costs above inflation".

Prices and capital

Inflation

• Base case: 2.5% a year for everything the Town buys other than staff.
• If it is wrong: 3.5% a year, −$8.6M.
• More: Utilities: in step with prices since 2019.
• Change it: "Inflation".

Capital

• Base case: The Select Board's policy from FY2028: 6.6% of the prior year's net revenue from revenue (4.5% for debt, 2.1% in cash), plus free cash. Debt follows the Town's six-year capital table through FY2032, and capital paid in cash makes up the rest of the 6.6%.
• Why: The FY2027 Financial Plan holds the policy at 6% for FY2026 and FY2027 only, for operating relief, and assumes the full 6.6% after that; its own table returns to 6.6% in FY2028. The base case applies the share to its own revenue, which counts the override passed after the table was written.
• If it is wrong: holding capital at the table's dollars, about 6% of this forecast's revenue, +$3.0M. The policy has been set aside before, as in FY2026 and FY2027.
• Change it: "Capital funding".

Reserves and other non-departmental lines

• Base case: By Town policy. The reserve fund is 1% of the prior year's net revenue. The other general lines (the stabilization, liability and housing trust deposits, insurance and small administrative items) follow the Town's plan to FY2031 and then grow with net revenue; items the Town pays without appropriation, such as the MBTA and county assessments and the overlay, grow about 2.4% a year.
• Why: These lines follow written policies or state law rather than a trend.

Growth

New growth rate

• Base case: New growth adds 1.23% of the levy a year, Brookline's rate over FY2022 to FY2026.
• Why: It is the recent record. Over FY2005 to FY2025 it averaged 1.37%; the fastest-growing fifth of Massachusetts towns reach 2.03%.
• If it is wrong: growth at 2.03%, half homes and half commercial, +$23.3M; led by commercial space, +$46.1M; led by new homes, +$5.6M.
• Change it: "Property tax new growth" and "Where added growth comes from".

New growth in FY2027 and FY2028

• Base case: FY2027 at the Town's budgeted $2.5 million, the figure in the adopted budget; FY2028 at the trend rate, as in every later year.
• Why: Building permits show new growth about two years ahead, and the 2024 and 2025 permits were the lowest since at least 2018. They point to real estate new growth of about $190 million in FY2027, close to the base case's about $195 million, and about $195 million in FY2028, against about $310 million. The base case keeps the trend for FY2028 until the state certifies FY2027's figure this fall.
• If it is wrong: FY2028 would have a shortfall of about $0.5 million instead of $0.8 million of room, and about $1.7 million a year less by FY2040, since each year's levy builds on the last.
• Change it: "New growth in FY2027 and FY2028" in the Excel workbook; the Explorer does not have this setting.

New homes built

• Base case: New homes at the recent pace, about 105 a year, about a sixth of new growth.
• If it is wrong: half as many, −$6.9M; none, with renovation, commercial space and business equipment at their pace, −$13.7M. That counts the tax and the Town costs of those homes but leaves the enrollment forecast as it is, since it is taken to reflect the recent pace of building. That is an assumption: the enrollment forecast has no housing input, and its grade-to-grade ratios carry whatever children recent building brought.
• Change it: "New homes built".

Town cost of serving a new home

• Base case: $660 a year (FY2025 dollars) in Town department costs, plus benefits. About a third of it, about $230, is police and fire. The forecast leaves fire's part, about $125, out, because it counts fire in steps instead (see "Police and fire staffing" above); police keep their part, about $105.
• Why: Comparing all 351 Massachusetts cities and towns from FY2011 to FY2021, those that added more homes spent about $660 more on non-school departments for each added home. The consultant RKG's average-cost figure for Brookline is $1,109; the low end of the estimates is $230.
• If it is wrong: $1,109, the consultant RKG's figure for Brookline, which has police and fire growing with every home instead of a company at a time: −$2.0M in FY2040 and −$13.0M over FY2026 to FY2040, mostly the higher charge for each home; $230, +$1.2M.
• Change it: "Town cost of serving a new home".

Students from new homes

• Base case: 0.244 school-age children per new home, each one staffed in full at the forecast's cost of an added student, $16,975 in FY2027.
• Why: 0.244 is a standard per-bedroom schedule applied to the mix of homes Brookline builds; the Town's count for ten Brookline buildings opened since 2015 runs from 0.03 to 0.56 and averages about 0.21, or about 0.13 in the five with the smallest homes.
• Alternative: With enrollment falling, children from new homes can fill seats the schools already have, at about $7,760 each for supplies and special education rather than new staff. New homes and Brookline's budget works through that case.
• If it is wrong: children from new homes filling empty seats until enrollment is back to the FY2027 level, at today's growth rate, +$0.8M; growth led by new homes at 2.03% with filled seats, +$18.1M, against +$5.6M with new staff.
• More: Students from new homes.
• Change it: "Students from added homes".

Planning and permitting

• Base case: When growth runs faster than the recent rate, Planning and building-inspection staff are added in proportion to the extra construction, from $2.6 million a year in the FY2027 budget. Slower growth does not cut them.
• Why: Permit fees already grow with construction on the revenue side; the staff who review it should too.

Chestnut Hill redevelopment

• Base case: The special permit filed in 2026 for the plan the spring 2026 rezoning anticipated: 384 homes, a 187-room hotel, about 41,000 square feet of retail and 94,000 of medical office, counted on top of trend growth. Its property tax enters the levy a third at a time from FY2032 to FY2034, at each year's value, and then grows with the levy; its hotel rooms tax and car excise grow 1.8% a year, like other local receipts.
• If it is wrong: counted inside trend growth instead, −$6.2M; at the consultant RKG's estimate, −$3.1M.
• Change it: "Chestnut Hill redevelopment".

Revenue

State aid

• Base case: Grows 2.5% a year from the adopted FY2027 figure, the rate in the Town's own long-range plan.
• Why: A conservative choice. State aid grew 3.5% a year from FY2015 to FY2025, but that includes a one-year rise of 8.6% in FY2019; without it the rate was about 3.0%. Since FY2019 it has grown 2.0% a year, and over twenty years about 2%, with a fall after 2008.
• Change it: "State aid growth".

Local receipts

• Base case: Fees, excise and other local receipts grow 1.8% a year, their FY2019 to FY2025 record with investment income set aside. Meals tax and building permits are forecast separately, and investment income is held at the Town's FY2027 budget of about $0.9 million a year.
• Why: Interest on the Town's cash rose from $1.3 million in FY2019 to $5.1 million in FY2025 as interest rates rose, nearly half the growth in all local receipts over those years. Part of that cash was federal pandemic aid waiting to be spent, which runs out in December 2026, and the Town's long-range plan expects interest income to fall as it is spent. Counted in the trend, it would carry a peak year forward; with it, receipts grew 3.3% a year.
• Change it: "Local receipts growth".

Free cash

• Base case: Each year's surplus comes back as free cash two years later, as it does in practice, and about 80% of it is appropriated. Following the Town's free cash policy, what is left after the capital and reserve uses already counted goes to capital and one-time uses, not the operating budget, so it is not counted as room.
• Why: It is the policy. The Town has at times used this money for operating costs; the base case does not.

Starting point and scope

Starting point

• Base case: The FY2027 budget Town Meeting adopted in May 2026, with the override. Each part of the budget moves with its own drivers from FY2028.
• Why: The years already budgeted are known. As a check, the forecast reproduces the adopted, balanced budgets, and its method, run over FY2011 to FY2025 with what actually happened to staffing, enrollment and prices, came within about 4% of Town department spending and 6% for schools.

Votes

• Base case: The May 2026 override phases in over FY2027 to FY2029. No further overrides or debt exclusions.
• Change it: "A further override" and "First year of that override" add one.

Debt exclusions

• Base case: Set aside on both sides. A debt exclusion is tax that matches specific building debt dollar for dollar, so leaving both in would add the same amount to revenue and spending without changing the room. Existing payments stay about level through FY2040.

Go deeper

• The base case to 2040, in full: If costs and revenue keep behaving as they have, where is Brookline's budget headed?
• Will Brookline need another override?: When would Brookline need another override, and what would prevent it?
• Forecast Explorer: What happens to the budget under your own assumptions?
• Data and sources: Where does each number come from?

Last updated 2026-10-26. What has changed.