Andover, Hebron, Marlborough & RHAM · FY 2013 – FY 2023 · What ten years of federal school-finance data say about rising per-pupil costs
A personal report by Scott Sauyet · scott@sauyet.com · Not an official town document
Every dollar figure in this report comes from one source: the school district finance survey (F-33) collected annually by the U.S. Census Bureau and published jointly with the National Center for Education Statistics, using the same definitions for every district in every year. Spending is current expenditure for elementary and secondary programs on the NCES basis, which includes Connecticut’s state-paid teacher retirement contributions, so levels here run higher than the state’s NCEP figures and the two should never be compared directly. Enrollment is F-33 fall membership, the October count, which matches the state’s PSIS-derived counts exactly. Inflation is CPI-U averaged over each July-to-June fiscal year. The most recent year published is FY 2023 (school year 2022-23). Figures for 2021-22 and 2022-23 were harmonized from Census files onto the NCES basis and verified to reconcile exactly, to the dollar, against the NCES series in the overlapping year. Town adjacency is computed from Census TIGER 2023 boundaries with mapped water removed, so towns count as neighbors only where they share dry ground. Group figures are enrollment-weighted: combined spending divided by combined students.
Introduction
When school budgets are debated in Andover, Hebron, and Marlborough, one number does most of the talking: cost per pupil. It is a genuinely useful number, and this report uses it throughout. But a ratio has two parts, and a per-pupil figure can rise because spending grew, because enrollment shrank, or both. Which one is doing the work matters enormously for what, if anything, a town should do about it.
This report separates the two. For each place examined, one chart shows four lines over the ten years from FY 2013 to FY 2023: per-pupil cost, total spending, enrollment, and inflation, all as cumulative change from the same starting year. Reading them together answers the question the per-pupil number alone cannot: are our schools getting more expensive, or are they serving fewer students?
The report starts with the three towns and their shared regional district taken as one system, then examines each of the four districts separately. It then widens the lens three times, to progressively larger contiguous blocks of surrounding towns, and closes with two reference groups: the 45 elementary-only districts that are Andover’s structural peers, and the state as a whole.
Methodology
Three methodological points shape everything that follows. First, the spending basis includes costs that never appear in a local budget, chiefly the state’s payments into the teacher retirement system on districts’ behalf. This is applied identically to every district in every year, so trends and comparisons are sound, but the dollar levels are not the town appropriation and are not comparable to NCEP. Second, all group figures weight by enrollment: they are the combined spending of the group’s districts divided by their combined students, so larger districts matter more, exactly as they do in the underlying totals. Third, regional school districts are separate entities from their member towns’ local districts. A Hebron seventh grader is counted in RHAM’s enrollment and nowhere else, so sums across a town-and-region group count every student and every dollar exactly once.
You do not have to take this report’s groupings as given. An interactive companion generates the same map, chart, and table for any town, district, or region you assemble yourself — same data, same rendering code — with a link you can share for whatever you find.
The Region
Andover, Hebron, and Marlborough run three local elementary districts and share one regional district, RHAM, for grades 7 through 12. RHAM’s territory is exactly the three towns, so the four districts together are the complete public school system of the region, and the natural place to start.
The combined picture over ten years is stark. Enrollment fell 30 percent, from 3,656 students to 2,555. Total spending rose 17 percent, which is well below the 29.5 percent rise in consumer prices over the same window: in real dollars, the region’s residents paid meaningfully less for their schools in 2022-23 than they did in 2012-13. Per-pupil cost nonetheless rose 68 percent. That is the arithmetic of decline: a real-terms spending cut, divided by a much faster enrollment decline, produces a per-pupil number that looks like runaway growth.
The chart makes the mechanism visible. The gold spending line runs below the purple inflation line for the entire decade. The blue enrollment line sinks steadily to 30 percent below its starting point. The green per-pupil line is simply the widening gap between gold and blue. Nothing in this picture shows a school system spending its way to higher costs; it shows a fixed-cost system absorbing the loss of nearly a third of its students.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 3,656 | $52,806,000 | $14,444 |
| 2014 | 2013-14 | 3,506 | $54,750,000 | $15,616 |
| 2015 | 2014-15 | 3,308 | $55,148,000 | $16,671 |
| 2016 | 2015-16 | 3,126 | $55,389,000 | $17,719 |
| 2017 | 2016-17 | 3,004 | $55,362,000 | $18,429 |
| 2018 | 2017-18 | 2,876 | $56,738,000 | $19,728 |
| 2019 | 2018-19 | 2,738 | $57,073,000 | $20,845 |
| 2020 | 2019-20 | 2,692 | $56,924,000 | $21,146 |
| 2021 | 2020-21 | 2,526 | $57,693,000 | $22,840 |
| 2022 | 2021-22 | 2,515 | $60,355,000 | $23,998 |
| 2023 | 2022-23 | 2,555 | $61,877,000 | $24,218 |
The Towns
Andover Elementary School shows the sharpest contrast in this report between the per-pupil line and everything underneath it. Over ten years its total spending grew 5.7 percent, roughly half a percent per year, against 29.5 percent inflation. In real dollars, AES’s budget shrank by about 18 percent. Enrollment fell 37 percent, from 314 students to 197. Per-pupil cost rose 68 percent, and every point of that rise above the spending line is the shrinking denominator, not new spending.
A town whose school budget trailed inflation for a decade does not have a spending problem. It has a small-district arithmetic problem: classrooms, buildings, and core staff do not scale down smoothly with each departing student, so the cost of the remaining system is spread over fewer and fewer children.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 314 | $4,705,000 | $14,984 |
| 2014 | 2013-14 | 298 | $5,054,000 | $16,960 |
| 2015 | 2014-15 | 275 | $4,922,000 | $17,898 |
| 2016 | 2015-16 | 250 | $4,962,000 | $19,848 |
| 2017 | 2016-17 | 225 | $5,047,000 | $22,431 |
| 2018 | 2017-18 | 204 | $5,024,000 | $24,627 |
| 2019 | 2018-19 | 191 | $5,097,000 | $26,686 |
| 2020 | 2019-20 | 195 | $5,001,000 | $25,646 |
| 2021 | 2020-21 | 173 | $4,766,000 | $27,549 |
| 2022 | 2021-22 | 191 | $4,888,000 | $25,592 |
| 2023 | 2022-23 | 197 | $4,975,000 | $25,254 |
The Towns
Hebron’s elementary district tells nearly the same story at three times the size. Total spending grew 4.1 percent over the decade, the most restrained spending line in the entire report, and far below inflation. Enrollment fell 26 percent, from 936 to 696. Per-pupil cost rose 40 percent, the smallest per-pupil increase of any group examined here, precisely because Hebron’s enrollment decline, while severe, was the shallowest of the three towns.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 936 | $14,717,000 | $15,723 |
| 2014 | 2013-14 | 867 | $14,047,000 | $16,202 |
| 2015 | 2014-15 | 785 | $14,291,000 | $18,205 |
| 2016 | 2015-16 | 730 | $13,527,000 | $18,530 |
| 2017 | 2016-17 | 705 | $13,515,000 | $19,170 |
| 2018 | 2017-18 | 701 | $13,622,000 | $19,432 |
| 2019 | 2018-19 | 663 | $13,600,000 | $20,513 |
| 2020 | 2019-20 | 658 | $13,700,000 | $20,821 |
| 2021 | 2020-21 | 636 | $14,227,000 | $22,369 |
| 2022 | 2021-22 | 679 | $14,984,000 | $22,068 |
| 2023 | 2022-23 | 696 | $15,320,000 | $22,011 |
The Towns
Marlborough sits between its neighbors: spending up 19.7 percent over ten years, still ten points below inflation, with enrollment down 29 percent, from 639 students to 456. Per-pupil cost rose 68 percent. As in Andover, the per-pupil line and the spending line diverge almost entirely because of the denominator.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 639 | $8,278,000 | $12,955 |
| 2014 | 2013-14 | 605 | $8,493,000 | $14,038 |
| 2015 | 2014-15 | 599 | $8,488,000 | $14,170 |
| 2016 | 2015-16 | 556 | $8,815,000 | $15,854 |
| 2017 | 2016-17 | 535 | $8,548,000 | $15,978 |
| 2018 | 2017-18 | 491 | $8,822,000 | $17,967 |
| 2019 | 2018-19 | 449 | $8,857,000 | $19,726 |
| 2020 | 2019-20 | 448 | $8,991,000 | $20,069 |
| 2021 | 2020-21 | 441 | $9,558,000 | $21,673 |
| 2022 | 2021-22 | 447 | $9,794,000 | $21,911 |
| 2023 | 2022-23 | 456 | $9,910,000 | $21,732 |
The Regional District
RHAM, the three towns’ shared district for grades 7 through 12, shows the region’s largest numbers on both sides of the ratio. Spending grew 26 percent over the decade, modestly below inflation. Enrollment fell 32 percent, from 1,767 students to 1,206, as the smaller elementary cohorts of the 2010s reached the upper grades. Per-pupil cost rose 85 percent, the steepest line in this report, and the decomposition shows why: RHAM combines the region’s deepest sustained enrollment slide with the fixed costs of operating a comprehensive middle and high school.
The elementary declines of the early 2010s arrive at RHAM’s door six years later, which means RHAM’s enrollment trough is partly still ahead even as the towns’ elementary counts have begun to stabilize.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 1,767 | $25,106,000 | $14,208 |
| 2014 | 2013-14 | 1,736 | $27,156,000 | $15,643 |
| 2015 | 2014-15 | 1,649 | $27,447,000 | $16,645 |
| 2016 | 2015-16 | 1,590 | $28,085,000 | $17,664 |
| 2017 | 2016-17 | 1,539 | $28,252,000 | $18,357 |
| 2018 | 2017-18 | 1,480 | $29,270,000 | $19,777 |
| 2019 | 2018-19 | 1,435 | $29,519,000 | $20,571 |
| 2020 | 2019-20 | 1,391 | $29,232,000 | $21,015 |
| 2021 | 2020-21 | 1,276 | $29,142,000 | $22,839 |
| 2022 | 2021-22 | 1,198 | $30,689,000 | $25,617 |
| 2023 | 2022-23 | 1,206 | $31,672,000 | $26,262 |
Wider Context
The remaining sections widen the lens for context. Block 1 adds every town sharing a land boundary with the core: Bolton, Colchester, Columbia, Coventry, East Hampton, Glastonbury, and Lebanon, eleven districts in all. On the maps that follow, green fills are districts in the aggregate, darkest in the core and one shade lighter with each added ring; blue-gray fills are towns counted through a regional district in the aggregate; hatched towns are named in the section and excluded from the figures.
The pattern softens but does not change. Combined spending grew 28.7 percent, still slightly below inflation: the entire contiguous eleven-district region cost marginally less in real dollars in 2022-23 than a decade earlier. Enrollment fell 17 percent. Per-pupil cost rose 55 percent.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 19,363 | $297,342,000 | $15,356 |
| 2014 | 2013-14 | 18,665 | $313,153,000 | $16,778 |
| 2015 | 2014-15 | 17,932 | $316,136,000 | $17,630 |
| 2016 | 2015-16 | 17,544 | $320,129,000 | $18,247 |
| 2017 | 2016-17 | 17,186 | $323,693,000 | $18,835 |
| 2018 | 2017-18 | 17,001 | $340,400,000 | $20,022 |
| 2019 | 2018-19 | 16,632 | $344,465,000 | $20,711 |
| 2020 | 2019-20 | 16,484 | $344,697,000 | $20,911 |
| 2021 | 2020-21 | 15,951 | $354,186,000 | $22,205 |
| 2022 | 2021-22 | 15,950 | $374,495,000 | $23,479 |
| 2023 | 2022-23 | 16,036 | $382,689,000 | $23,864 |
Wider Context
Block 2 extends to every town bordering block 1, twenty-four districts enrolling about 45,000 students. Spending growth (34.9 percent) now edges above inflation and the enrollment decline moderates to 11 percent. The per-pupil line, at 52 percent, barely moves: across every widening of the lens, per-pupil growth stays in a narrow band while its composition shifts from enrollment loss toward spending growth.
One town is hatched on the map and absent from the figures: Haddam, which runs no local district. Its schools belong to Region 17, which cannot join the block until its other member town, Killingworth, arrives in block 3.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 51,056 | $821,856,000 | $16,097 |
| 2014 | 2013-14 | 50,105 | $875,274,000 | $17,469 |
| 2015 | 2014-15 | 48,777 | $893,299,000 | $18,314 |
| 2016 | 2015-16 | 48,099 | $912,254,000 | $18,966 |
| 2017 | 2016-17 | 47,449 | $923,426,000 | $19,461 |
| 2018 | 2017-18 | 47,120 | $953,055,000 | $20,226 |
| 2019 | 2018-19 | 46,713 | $975,917,000 | $20,892 |
| 2020 | 2019-20 | 46,493 | $979,316,000 | $21,064 |
| 2021 | 2020-21 | 44,958 | $1,003,261,000 | $22,316 |
| 2022 | 2021-22 | 44,877 | $1,077,819,000 | $24,017 |
| 2023 | 2022-23 | 45,353 | $1,108,512,000 | $24,442 |
Notes: Excluded from the aggregate: Haddam (no local district; member of Region 17 (Haddam-Killingworth)).
Wider Context
Block 3 reaches roughly 96,000 students across forty-three districts, including two more regional districts (17 and 19) whose member towns fall entirely inside the block. The figures are now close to the statewide pattern: spending up 31.7 percent, enrollment down 12.1 percent, per-pupil up 49.8 percent.
Growing the block heals the Haddam gap: with Killingworth now inside, Region 17 joins, and both of its towns are fully counted through it (they render in blue-gray rather than hatching). Two towns remain hatched and outside the figures, Durham and Lyme, because their regional partners, Middlefield and Old Lyme, lie beyond every ring.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 109,093 | $1,810,086,000 | $16,592 |
| 2014 | 2013-14 | 107,169 | $1,918,291,000 | $17,900 |
| 2015 | 2014-15 | 105,512 | $1,957,397,000 | $18,551 |
| 2016 | 2015-16 | 103,818 | $1,988,823,000 | $19,157 |
| 2017 | 2016-17 | 102,462 | $1,990,713,000 | $19,429 |
| 2018 | 2017-18 | 101,254 | $2,037,840,000 | $20,126 |
| 2019 | 2018-19 | 100,267 | $2,066,505,000 | $20,610 |
| 2020 | 2019-20 | 98,840 | $2,063,893,000 | $20,881 |
| 2021 | 2020-21 | 95,268 | $2,119,183,000 | $22,244 |
| 2022 | 2021-22 | 94,901 | $2,294,380,000 | $24,177 |
| 2023 | 2022-23 | 95,921 | $2,383,959,000 | $24,853 |
Notes: Haddam has no local district but is fully represented through Region 17 (Haddam-Killingworth), included above; Killingworth has no local district but is fully represented through Region 17 (Haddam-Killingworth), included above. Excluded from the aggregate: Durham (no local district; member of Region 13 (Durham-Middlefield)); Lyme (no local district; member of Region 18 (Lyme-Old Lyme)).
Peer Districts
Andover, Hebron, and Marlborough belong to a structural peer group: the 45 Connecticut districts that, like them, operate elementary schools only and send their older students to regional or receiving districts. These are overwhelmingly small towns, concentrated in the state’s northwest and northeast corners. As a group they look like a milder version of the core: spending up 22.1 percent (seven points below inflation), enrollment down 17.1 percent, per-pupil up 47.3 percent. Real spending restraint amid enrollment decline is not a local anomaly; it is the normal condition of Connecticut’s small elementary districts.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 20,383 | $361,781,000 | $17,749 |
| 2014 | 2013-14 | 19,484 | $371,918,000 | $19,088 |
| 2015 | 2014-15 | 18,811 | $375,178,000 | $19,945 |
| 2016 | 2015-16 | 18,255 | $381,293,000 | $20,887 |
| 2017 | 2016-17 | 17,659 | $381,903,000 | $21,627 |
| 2018 | 2017-18 | 17,494 | $392,025,000 | $22,409 |
| 2019 | 2018-19 | 17,173 | $396,534,000 | $23,091 |
| 2020 | 2019-20 | 16,963 | $395,203,000 | $23,298 |
| 2021 | 2020-21 | 16,203 | $407,723,000 | $25,163 |
| 2022 | 2021-22 | 16,628 | $429,166,000 | $25,810 |
| 2023 | 2022-23 | 16,902 | $441,849,000 | $26,142 |
The Region
The state as a whole, held to a constant universe of 174 traditional local and regional districts, closes the sequence. Spending grew 33.6 percent, about four points above inflation. Enrollment fell 8.7 percent. Per-pupil cost rose 46.3 percent, meaning that even statewide, roughly a third of the decade’s per-pupil growth is the shrinking denominator rather than new spending. And the timing matters: essentially all of the state’s real spending growth occurred before FY 2019; from FY 2019 through FY 2023, the spending and inflation lines rise in near lockstep.
| Fiscal year | School year | Students | Total current spending | Per pupil |
|---|---|---|---|---|
| 2013 | 2012-13 | 517,812 | $8,529,875,000 | $16,473 |
| 2014 | 2013-14 | 511,082 | $8,979,397,000 | $17,569 |
| 2015 | 2014-15 | 505,366 | $9,197,103,000 | $18,199 |
| 2016 | 2015-16 | 499,494 | $9,371,901,000 | $18,763 |
| 2017 | 2016-17 | 496,074 | $9,487,857,000 | $19,126 |
| 2018 | 2017-18 | 491,590 | $9,792,860,000 | $19,921 |
| 2019 | 2018-19 | 486,809 | $9,999,141,000 | $20,540 |
| 2020 | 2019-20 | 483,151 | $9,925,529,000 | $20,543 |
| 2021 | 2020-21 | 468,502 | $10,249,922,000 | $21,878 |
| 2022 | 2021-22 | 468,926 | $11,020,709,000 | $23,502 |
| 2023 | 2022-23 | 472,943 | $11,394,627,000 | $24,093 |
At a glance
| Group | Per-pupil | Total spending | Enrollment | Spending vs. inflation (+29.5%) |
|---|---|---|---|---|
| Andover | +68.5% | +5.7% | −37.3% | 24 points below |
| Hebron | +40.0% | +4.1% | −25.6% | 25 points below |
| Marlborough | +67.8% | +19.7% | −28.6% | 10 points below |
| RHAM | +84.8% | +26.2% | −31.7% | 3 points below |
| The four combined | +67.7% | +17.2% | −30.1% | 12 points below |
| Block 1 | +55.4% | +28.7% | −17.2% | 1 point below |
| Block 2 | +51.8% | +34.9% | −11.2% | 5 points above |
| Block 3 | +49.8% | +31.7% | −12.1% | 2 points above |
| The 45 peers | +47.3% | +22.1% | −17.1% | 7 points below |
| Connecticut statewide | +46.3% | +33.6% | −8.7% | 4 points above |
Notes: All figures are cumulative change, FY 2013 to FY 2023, on the NCES F-33 basis described in the methodology. CPI-U rose 29.5 percent over the same window on July-to-June fiscal-year averages.
Takeaways
The region’s schools cost less, in real dollars, than they did a decade ago. Combined spending for Andover, Hebron, Marlborough, and RHAM grew 17 percent over ten years while prices rose 29.5 percent. Every one of the four districts, taken separately, also grew more slowly than inflation.
Rising per-pupil cost here is an enrollment story, not a spending story. The four districts lost 30 percent of their students. A ratio whose numerator shrank in real terms and whose denominator shrank far faster will rise; that is arithmetic, not extravagance.
The pattern weakens smoothly with distance. Each widening of the lens, from the core through three contiguous blocks to the state, trades enrollment decline for spending growth while per-pupil growth stays inside a narrow band, from 46 to 68 percent. The core towns are the extreme of a statewide gradient, not an exception to it.
Even statewide, the denominator does real work. Connecticut’s per-pupil cost rose 46 percent while spending rose 34 percent; the difference is an 8.7 percent enrollment decline. And the state’s real spending growth is concentrated before FY 2019; since then, spending has tracked inflation almost exactly.
Recent nominal jumps mostly reflect the inflation burst. The largest single-year spending increases in every group occur in FY 2022 and FY 2023, the same years CPI-U rose fastest. Nominal budget growth in those years is not evidence of accelerating real cost.
Small-district structure amplifies everything. The 45 elementary-only peers, as a group, show the same real restraint and deep enrollment loss as the core towns. Districts too small to shed fixed costs smoothly will always show the sharpest per-pupil arithmetic when students leave.
Enrollment has stopped falling, for now. All three towns and the combined region bottomed around FY 2021 and ticked up through FY 2023. RHAM, whose students arrive six years behind the elementary cohorts, likely has further decline ahead before any recovery reaches it.
Caveats
This report describes trends in one consistent federal data series. It does not answer several adjacent questions, and readers should not stretch it to cover them. It says nothing about educational quality or outcomes in any district. It does not explain why enrollment fell; that is a demographic question involving birth rates, housing, and migration, not a school-policy one. It excludes capital projects and debt service, which the F-33 tracks separately from current spending. Its dollar levels are not comparable to NCEP or to town appropriations, because the federal basis includes state-paid retirement contributions and other costs that never pass through a local budget. Federal pandemic relief (ESSER) is included in the spending of FY 2021 through FY 2023, so a portion of those years’ growth is one-time federal money rather than local burden. Fall membership counts include preschool where districts enroll it, and changes in preschool programs can move a small district’s denominator at the margins, though the enrollment declines shown here are far too large to be preschool artifacts. Finally, the most recent federal data is FY 2023; the two budget years since, including the referendums of 2025 and 2026, are outside its window.
To look past the ten groups presented here — different year windows, other towns, or a region of your own construction — use the interactive companion, which builds its charts from the same data files and rendering code as this report.
Documentation
Download
This report is available in three formats, all located alongside this page: