Pensions, health care and workers
Seventeen findings on what cities and towns pay for pensions and health insurance, what they owe for retiree health care, and the workers and schools behind those costs, for the state, each county and every town. Each figure is labelled as an annual cost (paid from one year’s budget) or a liability (a total owed over many years).
- Pension and health insurance as a share of residents' income
- When pension catch-up payments end, and what health costs take back
- Older pension debt and towns that lost population
- Teacher pensions the state pays for each town
- Growth in fixed costs against growth in state aid
- Who pays the property tax when many homes are seasonal
- Pension and retiree health costs inside regional school districts
- Retiree health care promised and owed
- How many public workers are near retirement
- The state health plan (GIC) against towns' own plans
- Births, school enrollment and regional district shares
- Can people who work in town afford to live there?
- Jobs located in town against working residents
- Wages against the cost of living
- Seasonal-worker visas and union elections
- Licensed child-care slots for young children
- Government as an employer, and state money in the local economy
When pension catch-up payments end, and what health costs take back
Most towns finish their pension catch-up payments (payments to make up for past underfunding) between FY2032 and FY2037, and the yearly payment that ends is about 9% of the property tax statewide. In about three towns in ten, projected growth in the yearly health insurance bill by that date is as large as the payment that ends. All figures are annual costs.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Yearly pension catch-up payment that ends, % of the property tax | annual cost | 9.0% | 4.0% (Norfolk) to 18% (Bristol) |
| Projected growth in health insurance by then, % of the property tax | annual cost | +7.3% | +2.2% (Dukes) to +16% (Plymouth) |
| Catch-up payment that ends, less projected health insurance growth, % of the property tax | annual cost | +1.7% | -2.5% (Plymouth) to +6.8% (Barnstable) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Yearly pension catch-up payment that ends, % of the property tax. ESTIMATE (pension unit shares, trend health, illustrative OPEB) per c1_release; county = sum of town dollars. Statewide figure: sum of town numerators / sum of town denominators.
- Projected growth in health insurance by then, % of the property tax. ESTIMATE (pension unit shares, trend health, illustrative OPEB) per c1_release; county = sum of town dollars. Statewide figure: sum of town numerators / sum of town denominators.
- Catch-up payment that ends, less projected health insurance growth, % of the property tax. ESTIMATE (pension unit shares, trend health, illustrative OPEB) per c1_release; county = sum of town dollars. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Older pension debt and towns that lost population
Towns that lost population since 1970 pay somewhat more per resident to catch up on past pension underfunding, but the link is weak and disappears from 1990 on. The pattern is clearer among the Gateway Cities, where older city retirement systems that were poorly funded carry the highest yearly payments whether or not population fell. Figures for towns in county retirement systems are allocated estimates.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Yearly pension catch-up payment per resident | annual cost | $298 | $155 (Franklin) to $562 (Suffolk) |
| Yearly pension catch-up payment, % of residents' income | annual cost | 0.53% | 0.28% (Dukes) to 1.00% (Hampden) |
| Population change since 1970 | +22% | -15% (Berkshire) to +291% (Nantucket, one town) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Yearly pension catch-up payment per resident. ESTIMATE for county-system members (unit share x system amortization); Barnstable County and Nantucket not in source. Statewide figure: sum of town numerators / sum of town denominators.
- Yearly pension catch-up payment, % of residents' income. ESTIMATE for county-system members (unit share x system amortization); Barnstable County and Nantucket not in source. Statewide figure: sum of town numerators / sum of town denominators.
- Population change since 1970. ESTIMATE for county-system members (unit share x system amortization); Barnstable County and Nantucket not in source. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Teacher pensions the state pays for each town
The state pays about $3.1 billion a year toward teacher pensions on behalf of cities, towns and school districts. The payment appears on no town's aid statement. Shared out in proportion to each district's teacher payroll, it is larger per pupil in towns with higher property wealth, the opposite of Chapter 70 school aid (the state's main school-aid formula). The link is weaker if the catch-up part is shared out by 2010 payroll instead. These are annual costs attributed to towns by arithmetic, not money any district receives.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Teacher pension the state pays, per pupil | annual cost | $3,238 | $2,644 (Hampden) to $4,821 (Dukes) |
| State school aid (Chapter 70, the main school-aid formula) per pupil | annual cost | $8,132 | $2,845 (Nantucket, one town) to $14,089 (Hampden) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Teacher pension the state pays, per pupil. ESTIMATE: FY26 MTRS appropriation x GASB68 payroll shares; regional members share one per-pupil figure. Statewide figure: sum of town numerators / sum of town denominators.
- State school aid (Chapter 70, the main school-aid formula) per pupil. ESTIMATE: FY26 MTRS appropriation x GASB68 payroll shares; regional members share one per-pupil figure. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Growth in fixed costs against growth in state aid
From FY2010 to FY2025, the yearly cost of employee benefits and insurance grew by $2.93 billion across all towns, while the yearly state aid towns keep after state charges grew by $2.34 billion. Health insurance for employees and retirees was $1.73 billion of that growth. Aid is mostly set aside for schools, so this compares totals; it does not show which costs the aid paid for.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Growth in pension, health and other insurance costs per $1 of growth in state aid after state charges, FY2010-25 | annual cost | 1.25 times | unstable where state aid barely grew or fell (Suffolk, Barnstable); see the county figures in the download |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Growth in pension, health and other insurance costs per $1 of growth in state aid after state charges, FY2010-25. actual: DLS Cherry Sheet / Schedule A; Chapter 70 goes to the regional district for regional-only towns. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Who pays the property tax when many homes are seasonal
Where many homes are seasonal, owners who live elsewhere pay a large share of the property tax: an estimated 38% to 65% in towns where 30% or more of homes are seasonal. Counting only the share of the property tax that year-round households pay, a town's yearly pension and health bills come to a much smaller figure per household there (for example $479 to $670 in Stockbridge, against $1,283 across all households). Statewide the effect is small, because only 31 towns have that many seasonal homes.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Homes that are seasonal | 3.7% | 0.4% (Norfolk) to 55% (Nantucket, one town) | |
| Pension and health insurance per household, all households | annual cost | $2,444 | $1,150 (Franklin) to $5,919 (Nantucket, one town) |
| Pension and health per household, counting only the year-round residents' share of the property tax (low estimate) | annual cost | $1,538 | $742 (Franklin) to $2,006 (Norfolk) |
| Same, high estimate | annual cost | $1,683 | $810 (Franklin) to $2,372 (Nantucket, one town) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Homes that are seasonal. ESTIMATE: seasonal share of levy (A owner address / B ACS units at value parity); fixed cost allocated pro rata to levy share. Statewide figure: sum of town numerators / sum of town denominators.
- Pension and health insurance per household, all households. ESTIMATE: seasonal share of levy (A owner address / B ACS units at value parity); fixed cost allocated pro rata to levy share. Statewide figure: sum of town numerators / sum of town denominators.
- Pension and health per household, counting only the year-round residents' share of the property tax (low estimate). ESTIMATE: seasonal share of levy (A owner address / B ACS units at value parity); fixed cost allocated pro rata to levy share. Statewide figure: sum of town numerators / sum of town denominators.
- Same, high estimate. ESTIMATE: seasonal share of levy (A owner address / B ACS units at value parity); fixed cost allocated pro rata to levy share. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Pension and retiree health costs inside regional school districts
Towns in a regional K-12 school district do not carry their school staff's insurance and pension costs on their own books; the district does and bills the member towns. Spread by enrollment, the district's yearly spending on insurance, retirement and other fixed charges adds, for the median such town, about 1.9 times the town's own pension and health bill. Shifts in enrollment shares also move part of the district's pension catch-up payments between member towns; the gross shift statewide is $0.68 million a year. Member shares are estimates from enrollment, not the assessments towns pay.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Regional school district's spending on insurance, retirement and other fixed charges for this town, % of the town's own pension and health | annual cost | 169% | 127% (Essex) to 322% (Hampshire) |
| Yearly change from shifting enrollment shares, $ per 1,000 residents (positive = pays less) | annual cost | — | — |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Regional school district's spending on insurance, retirement and other fixed charges for this town, % of the town's own pension and health. ESTIMATE: district shares by FY25 Chapter 70 foundation enrollment, not assessments; only towns with all pieces published. Statewide figure: sum of town numerators / sum of town denominators (towns in a regional K-12 district).
- Yearly change from shifting enrollment shares, $ per 1,000 residents (positive = pays less). ESTIMATE: shares from foundation enrollment; only 73 towns in K-12 regionals; shifts sum to zero within a district, not within a county. Statewide figure: gains and losses cancel within each district, so county and state figures are not meaningful.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Retiree health care promised and owed
Towns have promised retiree health care to current and future retirees and have saved little of it: in the 302 towns that report, the amount owed beyond savings is about $10,600 per household when every town is restated at one common interest rate (4.81%), and about 10% of the promise is saved as the towns report it. These are liabilities, totals owed over many years, not yearly costs.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Retiree health care owed beyond what is saved, per household (common 4.81% rate) | liability (total owed) | $10,617 | $4,932 (Franklin) to $25,364 (Nantucket, one town) |
| Share of the promise already saved, as the town reports it | 9.8% | 2.5% (Hampden) to 23% (Suffolk) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Retiree health care owed beyond what is saved, per household (common 4.81% rate). ESTIMATE: restatement by fitted duration; reported OPEB is actual (PERAC); 302 reporting towns only (49 not in PERAC list). Statewide figure: sum of town numerators / sum of town denominators.
- Share of the promise already saved, as the town reports it. ESTIMATE: restatement by fitted duration; reported OPEB is actual (PERAC); 302 reporting towns only (49 not in PERAC list). Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
How many public workers are near retirement
Public workers are older than the workforce as a whole, but the share aged 55 or older has been about flat since 2019. The age data are reported by county, so town figures are not shown.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Public workers (state and local) aged 55 or older, county rate (estimate) | 32% | 29% (Bristol) to 41% (Dukes) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Public workers (state and local) aged 55 or older, county rate (estimate). actual: county QWI public-sector 55+ share (carried through as county data); town 55+ counts are ESTIMATES = county rate x local-gov jobs. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
The state health plan (GIC) against towns' own plans
Towns that joined the state's Group Insurance Commission (GIC, the state-run health plan cities and towns can join) by July 2014 have lower health insurance costs than towns outside it on 6 of 7 measures, but the sample is not random and only some differences clear the usual statistical bar. Read this as worth testing properly, not as a finding. Town health insurance cost per resident is shown for every town.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Town health insurance cost per resident, FY2025 | annual cost | $572 | $254 (Hampshire) to $1,385 (Nantucket, one town) |
| Yearly growth in health insurance cost, FY2010-25 | +3.3% | +2.3% (Franklin) to +4.2% (Worcester) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Town health insurance cost per resident, FY2025. actual: DLS Schedule A (health_per_resident, growth); GIC/terms from a non-random 82-town sample; DLS health excludes regional-district school staff. Statewide figure: sum of town numerators / sum of town denominators.
- Yearly growth in health insurance cost, FY2010-25. actual: DLS Schedule A (health_per_resident, growth); GIC/terms from a non-random 82-town sample; DLS health excludes regional-district school staff. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Births, school enrollment and regional district shares
Public-school students living in Massachusetts towns fell 7% from FY2010 to FY2026, and births fell 8% from 2010 to 2023. Small hill towns fell most; some cities grew. A projection from births (an estimate, low confidence) shows a further 3.6% decline statewide by FY2031.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Public-school students living in town, change since FY2010 | -7.1% | -25% (Franklin) to +33% (Nantucket, one town) | |
| Projected change to FY2031 | -3.6% | -11% (Hampshire) to +14% (Suffolk) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Public-school students living in town, change since FY2010. actual: DESE School Attending Children, DPH births; projection is an ESTIMATE; county births change carried from source (Census components). Statewide figure: DESE statewide total 923,601 (FY2010) to 857,715 (FY2026); the town sum covers 342 towns with both years.
- Projected change to FY2031. actual: DESE School Attending Children, DPH births; projection is an ESTIMATE; county births change carried from source (Census components). Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Can people who work in town afford to live there?
In the median town, buying the median-priced home takes about 2.1 times the average yearly wage of a job located in that town (one household's needed income against one worker's average wage; 20% down, housing at 30% of income). Against the average teacher salary the median town needs 1.5 times. The table's statewide figures weight each town by its jobs, so they are lower than these town medians. The gap is large and uneven, and it tracks the price gap between where people work and where they live.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Income needed to buy the median home (20% down, housing at 30% of income), as a multiple of the average wage of jobs in town | 1.77 times | 1.26 times (Hampden) to 4.66 times (Nantucket, one town) | |
| Same, against the average teacher salary | 1.63 times | 0.94 times (Hampden) to 3.58 times (Nantucket, one town) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Income needed to buy the median home (20% down, housing at 30% of income), as a multiple of the average wage of jobs in town. ESTIMATE of income needed (PMMS 6.72%, FY26 tax rate, 30% rule); wages actual ES-202 2024; weights are jobs, so the county ratio is for the average job. Statewide figure: sum of town numerators / sum of town denominators.
- Same, against the average teacher salary. ESTIMATE of income needed (PMMS 6.72%, FY26 tax rate, 30% rule); wages actual ES-202 2024; weights are jobs, so the county ratio is for the average job. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Jobs located in town against working residents
Most towns have fewer jobs than working residents; job centers have more. Regional school districts put their staff on the payroll of the town that hosts the central office, so small host towns are adjusted, within a range, for the district's jobs.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Jobs in town per working resident | 1.09 | 0.72 (Plymouth) to 1.88 (Suffolk) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Jobs in town per working resident. ESTIMATE for the HQ adjustment; ES-202/LODES jobs actual. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Wages against the cost of living
In 287 of 351 towns, the average yearly wage of jobs located in town is below the living wage per earner for a family with two working adults and two children (MIT Living Wage Calculator). The averages include part-time jobs, so that count is somewhat overstated.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Average wage of jobs in town, % of a living wage (two working adults, two children) | 117% | 82% (Franklin) to 150% (Suffolk) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Average wage of jobs in town, % of a living wage (two working adults, two children). actual: MIT Living Wage (county, carried through), ES-202 2025 wages; income ACS 2020-24 nominal vs 2026 living wage; 2,080 hours per earner. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Seasonal-worker visas and union elections
Employers request about 6,460 seasonal-worker visa (H-2B) positions a year in Massachusetts, three-quarters of them in the most seasonal towns, mostly on Cape Cod and the Islands. Unions won 76% of union elections held in Massachusetts cities and towns since 2015. Town counts are too sparse to show by town, so county figures are used.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Seasonal-worker visa positions requested per 1,000 jobs | 1.8 | 0.1 (Franklin) to 120.0 (Nantucket, one town) | |
| Union win rate in National Labor Relations Board elections, 2015-26 | 76% | 54% (Hampden) to 88% (Bristol) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Seasonal-worker visa positions requested per 1,000 jobs. actual: NLRB tallies, DOL H-2B job orders (positions requested, not certified); jobs ES-202 2025. Statewide figure: sum of town numerators / sum of town denominators.
- Union win rate in National Labor Relations Board elections, 2015-26. actual: NLRB tallies, DOL H-2B job orders (positions requested, not certified); jobs ES-202 2025. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Licensed child-care slots for young children
There are about 0.41 licensed child-care slots per child under 5 statewide (0.55 counting home-based family care). By the common test of more than three children per licensed slot, 120 of 272 towns with 100 or more young children have too few slots. Public-school preschool is not counted, so supply is understated.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Licensed slots per child under 5 | 0.41 | 0.25 (Nantucket, one town) to 0.53 (Norfolk) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Licensed slots per child under 5. actual: EEC licensed capacity 2 Oct 2026 (programs placed by mailing-address city) and ACS 2020-24 under-5. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Government as an employer, and state money in the local economy
Government (federal, state and local) holds 12.2% of the state's jobs and 10.7% of its payroll. Where government jobs are concentrated is mostly not where state aid is concentrated. Gateway Cities receive state aid and teacher pension equal to 7.5% of residents' income, against 1.9% in other towns of 5,000 or more.
| Measure | Kind | Statewide | County range |
|---|---|---|---|
| Government jobs, % of all jobs in town | 12% | 9.9% (Middlesex) to 26% (Hampshire) | |
| State aid and state-paid teacher pension, % of residents' income | annual cost | 2.97% | 1.18% (Nantucket, one town) to 7.45% (Hampden) |
Every town and county: download town · county · state (CSV). Each town page shows its own figures.
For specialists: method and basis
- Government jobs, % of all jobs in town. ES-202 2024 jobs and wages; FY25 aid, FY26 teacher pension ESTIMATE; ACS 2020-24 income. Statewide figure: ES-202 statewide 2024 government share; the town sum differs because some jobs are not assigned to a town.
- State aid and state-paid teacher pension, % of residents' income. ES-202 2024 jobs and wages; FY25 aid, FY26 teacher pension ESTIMATE; ACS 2020-24 income. Statewide figure: sum of town numerators / sum of town denominators.
County and statewide figures are the sum of town numerators divided by the sum of town denominators. The joins behind these tables were checked independently in reviews r17 and r18, and these tables and pages in r19 (16 Nov 2026).
Massanomics Weekly Every Tuesday
Your city or town, in the numbers that matter.
- This week’s story, with where your city or town stands
- Economics, housing and the state money that reaches you
- Land, climate, history and politics
- Side-by-side findings that show how it all connects