Picture two ninth-grade classrooms on the first Monday of the school year. One is in Brattleboro, Vermont; the other is in Bakersfield, California. On paper, the difference in how many students each teacher answers for is close to double. Vermont reports roughly 11 students per teacher statewide. California reports closer to 21. That is one of the widest spreads in American public education, and it shows up year after year.
Below is what the enrollment-weighted numbers look like for the 2024-25 school year, drawn from NCES-reported enrollment and teacher full-time-equivalent counts across non-virtual public schools. I want to walk through the spread first, because the geography of it is striking. Then I want to do the harder part, which is asking what the number actually tells you, because that turns out to be a more interesting question than the ranking itself.
The Least Crowded States
The low end of the table is dominated by the Northeast, and it clusters tightly. Four states sit almost on top of one another at the bottom:
- Vermont · about 11.2 to 1
- New Hampshire · about 11.2 to 1
- Maine · about 11.3 to 1
- New York · about 11.3 to 1
- New Jersey · about 11.6 to 1
- North Dakota · about 11.6 to 1
- Massachusetts · about 11.8 to 1
- Connecticut · about 12.1 to 1
You can see the regional fingerprint immediately. Vermont, New Hampshire, Maine, New York, New Jersey, Massachusetts, and Connecticut form a near-continuous band. North Dakota is the outlier in the group, and it gets there a different way: a lot of small rural districts where a school of 80 students still needs a full slate of teachers, which mechanically pushes the ratio down.
The Northeastern pattern is more about money and norms. These are states with smaller districts, comparatively strong and stable school funding, and long-standing union staffing conventions that hold class loads down. A district that funds reading specialists, special-education staff, and intervention teachers will report a lower ratio than one that runs leaner, even if the kids are sitting in similarly sized rooms. Hold that thought, because it matters for the second half of this piece.
One footnote on the very bottom. Washington, DC comes in around 10.9 to 1, lower than any state. I am setting it aside from the state ranking because it is a single dense city school system, not a state, and comparing a city to North Dakota is not a clean comparison. It belongs in its own category.
The Most Crowded States
The high end leans heavily toward the fast-growing West and Sunbelt:
- California · about 21.5 to 1
- Utah · about 21.4 to 1
- Nevada · about 19.4 to 1
- Florida · about 18.0 to 1
- Idaho · about 17.9 to 1
- Oregon · about 17.8 to 1
- Washington · about 17.7 to 1
- Arizona · about 17.3 to 1
California and Utah sit at the top, essentially tied, and the rest of the crowded list is mostly Western: Nevada, Florida, Oregon, Washington, and Arizona. The common thread is growth. These are states where enrollment has climbed for years and teacher hiring has not kept pace, partly because of budget constraints and partly because of a thin teacher labor market in high-cost metros. Larger districts also play a role: a county-wide district of 300,000 students staffs to a tighter ratio than a New England town that runs its own small system.
From top to bottom, the spread is close to two to one. About 11 students per teacher in Vermont, about 21 in California. That is a real structural difference in how American schools are staffed, and it tracks geography and funding far more than it tracks any deliberate policy about ideal class loads.
A Word on the Data Before You Read Too Much Into It
These figures come from teacher full-time-equivalent counts that states report to NCES, and the reporting is not uniformly clean. A handful of states report teacher counts incompletely, which inflates or deflates their apparent ratio. So treat single-decimal gaps between adjacent states as noise rather than signal. Vermont at 11.2 and Maine at 11.3 are not meaningfully different; they are the same place on the map. I have left out states whose underlying counts look obviously partial, because a ratio built on an incomplete teacher count is worse than no number at all. The honest read of this table is the broad regional pattern, not the precise rank order within a tenth of a point.
Now the Part That Actually Matters: Ratio Is Not Class Size
Here is the distinction that gets lost in almost every conversation about these numbers. The student-teacher ratio is not the same thing as the size of the class your child sits in. They are different measurements, and the gap between them is large.
The ratio is a division problem at the school or state level: total students divided by total teachers. The denominator counts every teacher on the payroll. That includes reading and math specialists, interventionists, special-education teachers, part-time staff, and instructors who pull small groups out of the main room. Many of those teachers never lead a standard 25-student class. So a school can report a ratio of 14 to 1 while its average general-education classroom holds 26 children. The specialists are real, and they do real work, but they pull the ratio down without shrinking the room where most instruction happens.
This is why I am cautious when a district advertises its low student-teacher ratio to parents. It is a true number that answers a different question than the one parents are actually asking. If you want to know how many kids are in the room with your child during third-period algebra, the ratio will not tell you. Average class size will, and many states do not publish it cleanly. The concept of a good student-to-teacher ratio is worth understanding, but understanding it means knowing what it does and does not measure.
What the Research Actually Says
So does any of this predict outcomes for children? This is where the evidence gets more careful than the headlines.
The strongest causal evidence we have on class size comes from Tennessee's Project STAR, a randomized experiment run in the mid-to-late 1980s. STAR was unusual because it actually randomized students and teachers into small classes of about 13 to 17 or into regular classes of about 22 to 25, rather than just observing schools that happened to differ. The small early-grade classes produced measurable achievement gains. The effects were largest for low-income and minority students, and follow-up work found that some of those benefits persisted into later grades and beyond. STAR is the reason most researchers treat small early-grade classes as one of the few class-size interventions with solid causal backing.
The counterweight comes mainly from Eric Hanushek and a body of observational research arguing that, outside of those carefully designed early-grade reductions, the broader relationship between marginal ratio changes and student achievement is weak. Hanushek's long-running point is that shaving a student or two off an average class is expensive, since it means hiring more teachers and building more rooms, and the measured return per dollar tends to be small or inconsistent. When you look across the natural variation in class sizes nationwide rather than at a clean experiment, the signal mostly washes out.
The reasonable synthesis, the one I would actually stand behind, is this. Very small classes in the early grades, on the order of the STAR reductions, matter, and they matter most for the students who arrive with the least. Shaving a single point off an already-large high-school ratio probably does not move much, and it costs a great deal to do at scale. The two findings are not in conflict. They describe different parts of the curve. The early-grade end is where the evidence is strong; the marginal-adjustment-at-the-top end is where it is thin.
How to Read the Table
So what do you do with a state ranking of student-teacher ratios? Use it for what it is: a map of how states staff their schools, shaped by funding, district size, and growth, not a quality score. A low ratio in Vermont reflects small districts and strong funding. A high ratio in California reflects two decades of enrollment outrunning hiring. Neither number tells you what a specific classroom feels like, and neither one, on its own, predicts how a given child will do.
If you are comparing schools for your own family, push past the ratio. Ask the school directly about average class size in the grades that matter to you, especially the early grades where the evidence says size genuinely counts. Ask how many of those counted teachers actually lead full classes versus pulling small groups. The ratio is a starting point for a question, not the answer to one.
Data note: ratios are enrollment-weighted from NCES SY 2024-25 reported enrollment and teacher FTE for non-virtual public schools, via allk12.
Sources
NCES Digest of Education Statistics: Table 208.40, Public Teachers, Enrollment, and Pupil/Teacher Ratios by State
NCES Fast Facts: Teacher Characteristics and Trends (Pupil/Teacher Ratio vs. Class Size)
J-PAL (MIT): Project STAR, The Effects of Kindergarten Classroom on Earnings
Eric A. Hanushek (Stanford): The Evidence on Class Size
