The measure that didn't move

Part 4 of 4, a series on Vermont's schools statewide. Vermont's scores fell on every ruler that spans the decade. College-going fell with them. Chronic absence rose. The graduation rate moved three points. Since the class of 2020, a Vermont diploma certifies proficiency by definition — and on the current state test, 39% of students meet the math bar while 82% of seniors graduate. This story asks the question the series has been building to: if a diploma certifies learning, and measured learning fell this far, why didn't the graduation rate fall with it? And it asks who, in the system Vermont built, would have any incentive to let it.

This series laid out a decade of falling measures beside a graduation rate that barely moved, and ruled out the pandemic, measurable demographics, and special education. The second story reported how Vermont rebuilt its diploma around locally defined proficiency and changed its test three times. The third reported who steers: an agency answerable to the governor since 2013. This one puts the pieces together at the place they meet — the diploma.

1 · What a Vermont diploma promises

Since the class of 2020, graduating from a Vermont high school has meant, by rule, demonstrating proficiency. Not hours in a seat; mastery. But each district writes its own definition of proficient, and no external check stands between those definitions and the diploma. Vermont has never required an exit exam — and there it has company: the number of states requiring one fell from about twenty-seven at the peak to six for the class of 2026. A Vermont diploma is a local promise, differently written in every district, formally certifying the very thing the tests measure.

2 · The numbers side by side

On the current state test, 39% of Vermont students meet the math bar and 55% the English bar. The latest college-and-training figure is 46%. The graduation rate is 82%. One of these numbers is not like the others — and it is the one the diploma reports.

If a diploma certifies learning, and measured learning fell this far, why didn't the graduation rate fall with it? There are innocent readings. A diploma covers more than two tested subjects. NAEP samples fourth- and eighth-graders, not seniors. A senior class can fairly clear a local bar in a year its younger schoolmates test poorly. But there is also the reading the state itself has now put on the record: each district wrote its own definition of proficient, no external check stood between the definitions and the diploma, and a 2024 federal review found the rigor "uneven." Act 73 of 2025's statewide graduation standard, first binding on the class of 2031, is the state's own concession that the question was real.

3 · The number everyone is judged by

The graduation rate is not just a statistic — it is the one number federal accountability law grades every high school on. States must flag high schools whose four-year rate falls below 67% for comprehensive support. Districts publish the rate; boards answer for it; superintendents are measured against it. And the pressure works: the most rigorous national analysis found that accountability pressure "demonstrably caused" graduation-rate increases, especially in low-performing districts. Nationally, the machinery produced a spectacle in 2021-22: a record 87% of American students graduated in the same school year the national math scores posted their largest drop on record. High-school grade-point averages rose steadily through the same years that test scores fell — the testing companies' own data show it in the same students' records.

Vermont built a version of this machine with fewer brakes. The rate that schools answer for is public and consequential. The standard behind it is written district by district, with no external check. The agency overseeing both answers to a governor, and the third story in this series reported what that office's decade looked like. So the question extends naturally: in a system where every actor is judged on the graduation rate, and no actor outside the district checks what the diploma requires, what would keep the rate honest while the learning behind it fell? Nothing in the public record shows anyone leaning on a school to graduate students. What the record shows is a structure in which no one has to.

4 · What would answer the question

The question is answerable, and not by argument. The federal monitoring report behind the 2024 "uneven rigor" finding is a public record the state has so far only summarized — the full letter would show which indicator drew the finding. Each district's proficiency-based graduation requirements are public documents; setting them side by side would show how far the bar varies. And the proficiency profile of graduating classes — what share of diploma-earners met which standards — exists in district records. None of it is in the data the state publishes today. The class of 2031 will graduate under one statewide standard. The classes before it graduated under standards no one outside their own districts ever compared.

The scores fell. College-going fell with them. The diploma held still. Which of those numbers tells the truth about what a Vermont diploma certifies is the question the decade leaves.

What this doesn't say

No one in the public record is accused of inflating a graduation rate. The incentives the system creates are structural facts; what offsets them is the open question. The national graduation-rate rise is at least partly a real gain, not simply easier diplomas — the same analysis that documented the accountability pressure concluded the rise reflects "a real increase in human capital as well as some strategic behavior." A diploma legitimately certifies things no test measures. And the innocent readings in section 2 are genuinely live: the tested grades and the graduating grade are different students in different years. Any line from the decade's changes to the diploma's stillness is a question this series has posed, not a finding it has made.

Both tools carry the state-against-the-nation view

Explore the underlying data yourself: the Comparer · District Outliers.

---

Method and caveats: state proficiency figures are 2024-25 VTCAP, percent proficient or above, grades 3-9, weighted by students tested (English 54.6%, math 38.9%); the graduation rate is Vermont's four-year adjusted cohort rate (82% in 2024-25); college and training is the Annual Snapshot 16-month indicator (45.6%, 2024 snapshot, describing a class graduating about two years earlier). The tested grades and the graduating cohort are different students — the juxtaposition shows the system's numbers, not one cohort's. Vermont's no-exit-exam status and the national count (about twenty-seven states at the peak; six for the class of 2026) are per NCES and FairTest. The federal accountability threshold — high schools below 67% four-year graduation flagged for comprehensive support — is the Every Student Succeeds Act's identification requirement. The "demonstrably caused" finding and the real-gain conclusion are Harris (Brookings, 2020). The record-year juxtaposition is NCES Digest 219.46 (86.6% in 2021-22, rounded to 87% in the text) and the 2022 NAEP mathematics release; the grades-versus-scores divergence is ACT's longitudinal research (mean high-school GPA 3.22 to 3.39 across 2010–2021 while the ACT composite fell about a point). The 2024 federal monitoring finding is sourced to the Agency's own Act 73 report summaries; the underlying letter was not located and is named in the story as the record that would answer the question. Proficiency-based graduation requirements and graduate proficiency profiles are described as public or existing records; assembling them is reporting not yet done.

Sources: Vermont AOE assessment files, graduation series, Annual Snapshot; Vermont AOE Act 73 pages and the Act 73 statewide-graduation-requirements report (2025); NCES Digest 219.46; NCES/Nation's Report Card (2022 mathematics); NCES table dt12_199 and FairTest (exit exams); ESEA §1111 (comprehensive-support identification); Harris, Brookings (2020); ACT Research (grade inflation, 2022); Stanford Education Data Archive 2025.1.