Brookline fiscal model · expense driver tree
Brookline's General Fund spent $187.2 million in FY2011 and $364.7 million in FY2025. Every dollar of that $177.4 million increase in yearly spending can be assigned to one of nine causes.
How much more Brookline spent in FY2025 than in FY2011, by cause. Each figure is the difference in one year's spending, not a total added up over fourteen years. Hover or focus a row to see which budget lines it came from.
Inflation is the largest piece, at 42%. It is what it would have cost to keep the same staff doing the same work as pay and prices rose. Three more pieces account for most of the rest: debt service on school and other buildings, staffing levels (mainly more school staff per student), and paying down past pension underfunding. Growth in the town itself, new homes and new commercial space, is 5%.
of the growth came from Brookline's own decisions.
That covers how many staff it employs, what it builds and borrows for, how it prefunds retiree health, and how its pay and benefits compare with the benchmarks. Counting growth in homes and commercial space, which zoning and permitting shape, it is 41%. Most of the rest, 55%, was set outside Brookline or by obligations already incurred.
The lines between groups are judgment calls. Pay rates are bargained locally, so Brookline has some say over the part filed under inflation too; this page counts only the part of pay growth that differs from the national benchmark as a local decision. The pension shortfall built up over decades of earlier decisions, and the payment schedule is now fixed.
Each budget line starts from what it cost in FY2011. For the town departments and the schools, three questions are asked in order, and each answer is filed under one of the nine causes in the chart above:
Benefits and debt are handled separately because they don't behave like wages. Health insurance is split into premium inflation, more people covered, and a leftover. Pension contributions are split into the cost of pensions employees earn each year and the payment toward pensions already earned but not yet funded (Past pension shortfall). Debt service is categorized as Buildings.
The state's spending data for Brookline goes back to FY2005, but in FY2010 about $6 million moved between spending categories: the five town department categories fell 12.9% that year while total spending fell 2.3%. Comparing any year before FY2010 with any year after would mix that relabeling with real change, and it made General Government and Public Works look far more restrained than they were. FY2011 is the first full year on the new basis, and it is also the earliest year for which the Town's benefits detail and spending-by-type actuals are available in the Financial Plans used here.
The benchmarks. Over FY2011–25, national public-sector compensation rose 2.79% a year (+47% in total), Boston consumer prices 2.53% (+42%), and the public-sector benefits index 3.44% (+61%). BLS doesn't publish the pay index by state. The closest Massachusetts check is BLS average pay for local government workers, which grew 3.40% a year in Massachusetts against 3.65% nationally from 2014 to 2024. So the national benchmark doesn't understate Massachusetts pay growth; if anything it runs slightly ahead.
If Brookline had kept exactly the FY2011 staff and services, yearly spending would still have grown by about $75 million, because pay, prices and health premiums all rose. Schools carry the largest share because they are the largest budget line.
Non-personnel spending needed its own check, because the method assumes it grows with consumer prices. The Town's own actuals by type of expense (services, supplies, utilities, capital outlay) show it did. From FY2011 to FY2024 it grew 2.9% a year, or 2.5% without an "Other" line whose contents appear to have changed, against 2.5% for Boston prices. Personnel spending grew 4.1% a year over the same years, which is where most of the rest of this page comes from.
For the forecast. This piece depends on two rates: public-sector pay growth and health premium growth. It is the largest single piece, so those two assumptions carry more weight than any other.
Debt service rose from $9.5 million to $36.3 million a year. Most of the increase is debt voters approved for specific projects: the Devotion School (now Florida Ruffin Ridley) expansion (approved 2015), the high school expansion and renovation (2018), the Driscoll School expansion (2019), the Newbury College land purchase (2020), fire station renovations (2022) and the Pierce School (2023). Devotion and Driscoll were expansions approved during the enrollment rise shown further down.
Excluded debt is paid by a separate tax that voters approve for each project, outside the Proposition 2½ levy limit. So the $20.8 million grew the budget without taking room under the limit from operating spending.
The $6.0 million of routine borrowing does take room under the limit. It follows a standing Town policy: about 6% of the prior year's net revenue (6.6% today) goes to capital each year, split between projects paid in cash and debt service on projects financed within the levy. Within-levy debt service went from $8.2 million to $14.3 million, rising roughly with revenue as the policy intends. The capital plan's debt schedule lists the individual projects; they aren't itemized here.
For the forecast. Repayment schedules for existing debt are known in advance, so this is the easiest piece to project. The Town already projects excluded debt service at about $36–37 million a year from FY2028 on.
Brookline had 6.75 students per school staff member in FY2011 and 5.62 in FY2025. If staff had moved one-for-one with enrollment, the schools would have about 208 fewer positions today. At about $94,000 each in FY2025 salary, that gap is the $19.6 million. Health insurance and pensions for those positions are in the benefits lines.
The $94,000 is an average across all school staff, including paraprofessionals, aides and other roles paid less than teachers; the state reports Brookline's average teacher salary at $132,000 in FY2025. It comes from salaries being about 81% of school spending, with the rest going to special education tuition, transportation, supplies and similar costs that don't scale with staff.
Comparing only FY2011 and FY2025 hides how it happened, because enrollment rose and then fell:
Index, FY2011 = 100. Hover for values. Staff data begin FY2008.
From FY2011 to FY2019, enrollment rose 18.5%, to 7,855, and staff rose 34%. After COVID, enrollment fell 10.6% by FY2025 while staff fell 5.0%. Measured separately, about 150 of the positions beyond enrollment were added while enrollment was rising, and about 73 are positions kept after it fell.
Two measured changes run alongside this. Students with an IEP rose from 1,046 to 1,270 (+21%, against +6% for enrollment), from 15.8% to 18.1% of students. And per-pupil spending grew fastest in pupil services (5.9% a year FY2009–25) and guidance and counseling (4.6%), while teacher spending per pupil grew 2.7%. Neither establishes why the ratio changed; both show where the added spending went.
On the town side, departments added 48.9 positions. About 24 of them are what new homes and commercial space would call for at the model's rates, leaving about 25. Of the 48.9, 18.5 appear at a break between two data sources in FY2020/21 and may reflect a change in how positions were counted rather than hiring.
For the forecast: this is one of the most important assumptions. The school staffing ratio decides a large part of future spending. Holding it at 5.62 students per staff member, returning part way toward FY2019 levels, and letting staff follow enrollment are three different expense paths, and the forecast should show all three rather than pick one silently.
The Town's pension contribution rose from $14.0 million to $37.5 million a year. Part of that pays for pensions current employees earn each year, which actuaries call the normal cost. It rose from about $3.3 million to $7.3 million, and it is split across three causes: pay inflation ($1.3 million), more town staff ($0.3 million), and a higher cost per dollar of pay ($2.3 million), since the normal cost went from 5.48% to 7.71% of payroll as the retirement system's actuarial assumptions changed. That last piece is under Pay & costs vs inflation.
The rest pays down the gap between pensions already promised and the money set aside for them. That payment ends when the system reaches full funding, which the Town projects for 2030. After that the pension contribution falls to the normal cost. The Town's own long-range schedule redirects most of that room to retiree health (OPEB) starting in FY2031, at about $43 million that year.
For the forecast. This is set by a published schedule through 2030, then changes sharply. How the room freed in FY2031 is used is a policy choice the forecast has to represent explicitly.
Turnover with the same number of positions moves costs in both directions, and only one direction lasts past the employee's departure.
New homes add real costs, and they are small next to everything else. $9.5 million a year compares with $75.3 million for inflation, $26.8 million for debt service and $26.5 million for staffing levels. This page covers costs only; the tax revenue those homes bring is a separate question.
About 1,638 homes were added over FY2012–25, estimated from new-growth tax dollars; three independent counts agree on the order of magnitude. The fiscal model charges $1,371 per home for town services, taken from the Town's consultant RKG. Of that, $262 is benefits, debt and insurance, which this page counts in the benefits lines instead, so the town-department share used here is $1,109 per home, or $1.8 million.
RKG puts 88% of that $1,109 in police ($448) and fire ($532), with general government at $57, public works $48, library $19 and the rest under $10 each. That reflects RKG's judgment, line by line through the FY2025 budget, about which costs rise with households. Brookline's actual staffing went the other way: public safety positions fell by 5 over FY2011–25, while general government added 24.6, human services 17.8 (12.5 of them at the FY2020/21 data break), public works 8.0 and culture and recreation 3.5. The total added cost roughly matches the model's rate in the one window measured directly (FY2022–24, $1,349 per home), but not the mix.
At the model's 0.244 students per new home, those homes would bring about 400 students. Staffed at the FY2011 ratio and FY2025 pay, they account for $5.6 million. A Town and Schools count of ten developments of 8 or more units built since 2015 found between 0.03 and 0.56 enrolled students per unit in Fall 2025, a simple average of about 0.21, so the model's figure sits in the observed range. Net enrollment change over the window was also about 400, which is why Enrollment not from new homes nets to nearly zero. That match is a coincidence of endpoints: the path above moved much more than new housing could supply.
Non-residential new growth totalled $688 million over FY2012–25, but most of that is personal property, mainly utility equipment, which creates no service demand. In FY2026, the one year with a published split, personal property was 74% of it. At that split, about 350,000 square feet of commercial space was added, roughly 7% of today's stock of about 5.1 million, costing about $0.3 million a year in department services. Treating all non-residential growth as commercial space gives an upper bound of $1.2 million.
For the forecast. Because this piece is small, differences between growth scenarios will move expenses less than the pay, staffing and pension assumptions do.
The $4.0 million is the difference between what the state's Schedule A reports as fixed costs and what the Town's benefits detail lists. It likely includes property and liability insurance and differences in how the two sources classify items. It isn't itemized in either source.
This is the leftover after inflation and staffing. It holds step and grade raises, turnover, changes in the mix of roles, any gap between Brookline pay and the national benchmark, and, for schools, non-personnel costs growing faster than prices.
| Budget line | Growth | Inflation | Staffing | Above / below | FTE change |
|---|---|---|---|---|---|
| Schools | $69.0M | $35.7M | $25.0M | +$8.3M | +267.1 |
| Public Safety | $15.5M | $13.0M | −$0.5M | +$3.0M | −5.0 |
| Pensions earned each year | $4.0M | $1.3M | $0.3M | +$2.3M | n/a |
| Health insurance | $18.4M | $12.1M | $6.1M | +$0.2M | n/a |
| Culture & Recreation | $2.8M | $2.6M | $0.3M | −$0.1M | +3.5 |
| General Government | $5.2M | $4.3M | $2.4M | −$1.4M | +24.6 |
| Human Services | $0.8M | $0.9M | $1.6M | −$1.7M | +17.8 |
| Public Works | $3.1M | $5.4M | $0.7M | −$3.0M | +8.0 |
Schools ran $8.3 million above the benchmarks. Two things measured elsewhere fit that: the state reports Brookline's average teacher salary rising from $96,000 in FY2018 to $132,000 in FY2025, about 4.7% a year against 3.3% for the national wage benchmark (an average that also moves with the experience mix of the teachers employed), and special education and other non-personnel costs aren't limited to consumer-price growth. Public Safety is the one town department well above the benchmark, with headcount slightly down. Public Works came in $3.0 million below; about 39% of its budget is contracts, fleet and materials, where the consumer price benchmark fits less well. Health premiums tracked the national public-sector benefits index almost exactly.
The same $177.4 million, by cause and by budget line. Shading shows size. Reading down the columns, new homes and commercial space are a small entry in every budget line they touch.
Net effect: Staffing levels fell from $30.3 million to $26.5 million, New homes & commercial space from $11.2 million to $9.5 million, Past pension shortfall from $21.1 million to $19.6 million, and Pay & costs vs inflation rose from −$1.4 million to +$7.6 million.
In order of how much each could move the picture:
Every total on this page is checked in the script: the nine causes sum to the $177.4 million change to the dollar, each budget line sums to its own change, and the town department split reproduces the earlier Phase B result.