Finding 1 · Survives every peer group
Graduates near the top, college and training near the bottom
Washington Central gets students to a diploma at close to the best rate in the state — 89%, fifth of 43 systems. Then it loses them. In the latest measure, 43.1% enrolled in college or training within 16 months, against a close-peer median of 52.5% — behind every one of the thirteen peers reporting. (The state’s snapshot lags graduation: the ’23-24 edition covers the class of about ’21-22.)
The cliff came in two years, not ten
Share of graduates continuing to college or training within 16 months, by Annual Snapshot year
Vermont Annual Snapshot, college & career readiness. 35 of 46 systems report; each snapshot reflects the class that graduated about two school years earlier.
The decline is not a slow slide: the ’18-19 through ’20-21 snapshots plateaued between 58% and 66%, then fell by nearly thirty points in two years. Peers dipped in the same years and have taken most of it back — the median close peer is at 52.5% from a low of 43.5%. Washington Central sits at 43.1%, fifteen points below the bottom of its own old plateau.
Finding 2 · Survives every peer group
The decline is in the upper grades, not the lower ones
Every Vermont district is shrinking. Washington Central’s kindergarten decline (−20% over the decade) is ordinary — the statewide median is −15%. What is not ordinary is where the students go afterward: senior-class enrollment fell −32%, sixth-worst of all 46 systems, against a median of −10% — and all five of the systems that fared worse are supervisory unions where some towns tuition students out by design.
The leak is in the secondary grades, and it is accelerating. Elementary grade-to-grade retention is unchanged since the five town districts merged. All but one secondary transition weakened — and the jump from junior to senior year most of all: 91% of juniors returned as seniors before unification, 82% since.
The last two cohorts are worse than the average. The class of ’23-24 lost 19% of its juniors (108 → 87). The class of ’24-25 lost 33% (121 → 81) — forty students gone between junior and senior year.
Enrollment data alone can’t say where they went: Early College, transfer, and dropout look identical in a fall head count.
One explanation would soften the finding, and it deserves to be named. Vermont law lets a senior spend the year enrolled full-time at a college instead, tuition-free — and a student who takes that route is, by statute, struck from the district’s enrollment count and its pupil count for the year (16 V.S.A. § 946; districts record the exit as a transfer to a college program). A school where Early College is popular will show a smaller senior class than it actually produced. Statewide, 433 seniors — 8 percent of the class, a record — took that route in 2024-25 (AOE, March 2026). Typical uptake does not cover a 33 percent cohort loss; uptake well above the state’s could cover a real share of it. Whether Washington Central’s is, the public files don’t say.
If Early College is the explanation, it is not a neutral one. A senior class of 81 where 121 juniors stood is thinner for the students who stay, and the one randomized study to look at the question found early-college students took part in sports, arts, and other school activities less than their peers (Edmunds et al., 2026). Most Vermont Early College seats are at the Community College of Vermont, and the research on starting college at a two-year school points one way: in the most careful state-level study, students who intended a bachelor’s degree and started at a community college were about 15 percentage points less likely to finish one within nine years than similar students who started at four-year schools (Long & Kurlaender, 2009) — largely because too few transfer, and credits are lost when they do (Monaghan & Attewell, 2015). Washington State’s Running Start, the closest studied cousin of Vermont’s program, raised college attendance overall but lowered four-year attendance in the year after high school (Cowan & Goldhaber, 2015).
The strongest evidence in Early College’s favor — randomized lottery studies showing higher degree completion — studied a different animal: small schools built around the model from grade 9 up, not a single senior year at a community college. Even there the gains came through associate degrees, and ten years on the bachelor’s-degree advantage had washed out. Vermont has never evaluated its own version against a comparison group, and the state’s descriptive follow-ups cut both ways: nearly all Early College students finish the year, but roughly one in five enrolls in no college at all the following fall (VSC legislative report, 2024). So Early College cannot be assumed to lift a school’s college-continuation number. Washington Central’s fell.
What would settle it: the district’s own Early College and dual-enrollment counts by year, and where those students stand a year later — numbers the public files don’t carry. Forty students is a fifth of a class. Better data is needed to answer this one conclusively, and this report leaves it open.
Where each grade-to-grade transition changed since unification
Average cohort retention, pre-union (through ’18-19) vs. post-union (’20-21 on), percentage points
AOE enrollment collections; predecessor town districts summed onto a consistent footprint.
Finding 3 · How it got here matters more than where it is
From near-best attendance in Vermont to average
Washington Central’s chronic-absenteeism rate — 25.1% in ’23-24 — is indistinguishable from the statewide median. The problem is how it got there. In ’17-18 it had the fourth-best attendance of 44 reporting systems, within a point of the best. It held that advantage straight through the COVID year that broke everyone else. Then it surrendered all of it in a single year.
One year erased a decade of advantage
Share of students chronically absent (missing ≥ 10% of school days)
Vermont Education Dashboard, chronic absenteeism (both statuses). ’20-21 spans remote and hybrid attendance counting — read gently.
The initial recovery from the ’21-22 spike was quick — fourteen points in the first year. It has since stalled: in the latest year Washington Central improved −1.2 points while the median close peer improved −2.2. The district that went into the pandemic eleven points ahead of its peers is now nearly four behind them.
Finding 4 · Rewritten July 11 with the fuller comparison set
In science, behind the districts most like it — in math, no longer
Statewide, Washington Central's scores look strong — 11th of 46 in math, 16th of 44 in science. But the statewide table mixes in systems far poorer than Washington Central. Against the fifteen systems most like it, science is ordinary: 46.3% against a peer median of 47.9% (10th of 16).
Math is a different story now. As first published — against the ten peers the original sourcing covered — math trailed too: 45.1% against a 47.7% median. The expanded sourcing brings Vermont's supervisory unions into the comparison, and the systems genuinely most similar to Washington Central score lower in math than the wealthier districts that had stood in as peers. With the fuller set, math reads 45.1% against a peer median of 39.9% (5th of 16) — above the median at every peer-group size from ten to the full state.
The direction still bears watching: on the current test, math has slipped −1.7 points since ’22-23 while the median close peer held about flat (+0.2). Science moved the other way — up +4.0 points against a peer median of +1.4: behind on the level, gaining on the trend.
Math improved against the state after unification — and now holds up against peers too
Gap to the Vermont average in math proficiency, percentage points
AOE assessment files; gap to state computed within each test era. No adjusted view is shown for ’21-22 (test transition).
The improvement against the state since unification is real — math was 4.6 points below the Vermont average before the merger. Science has not made the same trip against the peers, and that is what remains of this finding. English is the exception, ahead on every comparison.
For balance
What holds up
Two things hold up on every comparison — close peers, the full state, and demographic expectation.
English proficiency shows no swing at all — 7 to 11 points above the state in every observed year since ’17-18, before and after the merger. It is a steady strength, not a recent one.
Multi-year shifts, single-year noise excluded
What changed since the five districts became one
Washington Central unified in 2019-20. Comparing the constituent-district era (through ’18-19, town districts summed) with the current one:
The last row bears on the others: the student body Washington Central serves today is measurably less affluent, relative to Vermont, than the one its constituent districts served. Holding test scores steady across that shift counts for something — and it is why the enrollment and postsecondary numbers, which did move, deserve the attention.
Provenance
Method & caveats
Every claim is tested against two groups. The fifteen closest peers — nearest to Washington Central by enrollment, share of low-income students (Act 127 direct certification), and median family income (Census ACS, town medians weighted across each system’s member towns — family income, not household income, so towns with many one-person households read correctly) — and all 45 other Vermont systems operating K-12 schools with current data, unified districts and supervisory unions alike. Findings labeled “survives every peer group” hold in both, and at every group size in between.
- Mt Abraham Unified
- Lamoille South°
- Addison Central
- Addison Northwest
- Windsor Southeast°
- Montpelier Roxbury
- Hartford
- Milton
- Rutland Northeast°
- Orleans Southwest°
- Caledonia Central°
- Franklin West°
- White River Valley°
- Colchester
- Mountain Views°
° marks a supervisory union of several member districts; some member towns there tuition students to schools outside the union, so retention and continuation figures partly reflect that structure. Six systems on the state’s supervisory roster are not compared here: Bennington Rutland, Grand Isle, and St. Johnsbury (schools through grade 8), Kingdom East (through grade 9), and Lincoln (through grade 6) operate no high school — their students tuition out, so K-12 measures don’t exist at their grain — and Norwich belongs to the Dresden interstate district, which reports through New Hampshire’s SAU 70. Caveats: postsecondary is reported by 35 of 46 systems, and the Clearinghouse match behind it misses apprenticeships, military service, and most sub-two-year certificate programs — systems sending graduates into the trades read low; district-grain spending is reported by 23; ’20-21 attendance spans remote/hybrid counting. District income is a family-weighted mean of town ACS family-income medians, not a true district median. Enrollment windows cross Act 46 reorganizations; predecessor districts are summed onto a consistent town footprint. Chronic absenteeism is pulled from the live state dashboard with both status counts, so rates use the dashboard’s own denominators; one system’s ’17-18 rate is reported near 100% and shows as an extreme improvement at the far left of the change strip — read that tick gently. The comparison pool and peer group widened on July 11, 2026, when the low-income data sourcing was expanded — see the editor’s note.
Every number regenerates from the same public-data pipeline as the interactive tools at hungermountaini.com. Check the work:
See where any district stands apart →
Editor's note, July 11, 2026. This story first ran on July 10 comparing Washington Central within a set of 30 Vermont K-12 systems and a close-peer group of ten — every system whose Act 127 low-income share the data pipeline could then source. On July 11 the sourcing was expanded to cover Vermont's supervisory unions alongside its unified districts, and several unions' shares now reflect their full membership. The comparison set is now 46 systems, and the peer group fifteen.
Three of the four findings — college continuation, the upper-grade enrollment decline, and the absenteeism deterioration — hold in the fuller data at every peer-group size, as do the strengths. The math comparison reads differently. As first published, Washington Central's math proficiency (45.1%) trailed its ten closest peers' median of 47.7%, sixth of eleven. The expanded set brings in the rural systems genuinely most similar to Washington Central, and against them its math stands above the peer median at every group size. Science remains behind. Finding 4 has been rewritten, and the numbers and charts throughout reflect the fuller comparison set. A same-day refinement replaced town household income with family income in the peer measure — household income was misreading towns with many one-person households, Montpelier most of all — which swapped three of the fifteen peers (Montpelier Roxbury moved closer, as neighbors should when the measure is fair) and moved no finding.
