When I became senior director of special education for Granville County Schools, a rural district north of Raleigh, the district's out-of-district budget was $0. The district serves about 6,400 students across a county of more than 500 square miles, and its students were educated by its own staff, in its own buildings. Federal data for 2024-25 still shows no students from the district in a separate day school.
Before and after Granville, I led special education in three Massachusetts districts north of Boston, each with half of Granville's enrollment or less, where out-of-district budgets ran past $1 million a year. Going back and forth between the two states, I worked under the same federal law and the same promise of the least restrictive environment, and the two pictures could hardly have looked more different. I've carried that contrast around for years without a good way to explain it, and this week the federal data in the Special Education Observatory gave it a shape.
What the map shows
The federal government asks every district where its students with IEPs spend the school day. The most common measure of inclusion is the share of students who are in general education classes for 80% or more of the day. Nationally, for school-age students (ages 5 in kindergarten through 21) in 2024-25, that share is 67.9%. Here's how a few states compare.
State | In general education 80% or more of the day | Rank of 51 | In separate day schools | Share of students with IEPs |
|---|---|---|---|---|
Alabama | 82.5% | 1 | 1.2% | 13.7% |
Vermont | 82.0% | 2 | 6.3% | 18.3% |
New Hampshire | 81.8% | 3 | 0.8% | 18.1% |
Mississippi | 78.9% | 8 | 1.0% | 15.7% |
Texas | 75.4% | 16 | 0.3% | 14.7% |
North Carolina | 71.0% | 25 | 1.0% | 13.1% |
Connecticut | 67.2% | 30 | 5.2% | 18.1% |
Massachusetts | 67.0% | 32 | 5.4% | 19.8% |
New York | 58.2% | 46 | 4.5% | 21.5% |
New Jersey | 44.9% | 51 | 6.1% | 17.4% |
United States | 67.9% | 2.5% | 15.3% |
Source: U.S. Department of Education, IDEA Section 618 educational environments and child count data, 2024-25, as compiled from district reports in the Special Education Observatory. "Separate day schools" counts separate day schools only; residential facilities, homebound or hospital, and correctional settings are counted separately. Ranks are among the 50 states and DC.
Three things surprised me
Southern states rank near the top. Alabama is first in the country. Mississippi is eighth and Texas is sixteenth. Massachusetts, a state that so often sits at the top of national education rankings, is thirty-second.
The pattern is more tangled than a regional story. Vermont and New Hampshire, two New England states, are second and third. Vermont is the one I keep coming back to. It ranks second for inclusion and also has the second-highest share of students in separate day schools in the country, at 6.3%. A state can include most students most of the day and still send more students to separate schools than almost anyone else.
The biggest gap is in separate day schools. Nationally, 2.5% of students with IEPs attend one. In Massachusetts it's 5.4%, in Connecticut 5.2% and in New Jersey 6.1%. In North Carolina it's 1.0%, and in Texas it's 0.3%. Those are the $0 and $1 million budgets I walked into, showing up in federal data: Massachusetts places students in separate day schools at about five times the rate of North Carolina.
Why it matters
Inclusion is one of the most researched practices in our field. A 2013 study by Cosier, Causton-Theoharis and Theoharis found that students with disabilities who spent more time in general education classes had higher reading and math achievement. A 2016 review of the research led by Thomas Hehir for Abt Associates found that students with disabilities generally did better in inclusive settings, and that their classmates without disabilities did as well or better. And in 2014, Kurth, Morningstar and Kozleski found that highly restrictive placements for students with the most significant disabilities had persisted over time, and that the setting a student was placed in varied widely by state.
Put those together and the uncomfortable conclusion is that a student's zip code shapes where that student learns, and possibly how much that student learns.
The questions I can't answer yet
I want to be honest that I don't know why the map looks this way. Here are the questions I'm asking, and I'd love your help with them.
Is it about who gets identified? Massachusetts identifies 19.8% of its students for special education, and Alabama identifies 13.7%. You might expect the state that identifies more students, many with milder needs, to have a higher inclusion rate. Across all 51, though, the relationship between a state's identification rate and its inclusion rate is weak. Are some states identifying students for more intensive services than they need? Are others missing students who need support? Researchers have found that under-identification is real: Morgan and colleagues reported in 2015 that students of color were less likely to be identified than otherwise similar white students.
Is it about what's nearby? Massachusetts passed its own special education law, Chapter 766, in 1972, three years before the federal law, and a network of approved private day schools and collaboratives grew up around it. When a specialized program exists twenty minutes away, it gets used. Around Granville, there were very few of those programs within reach, and the ones that existed often had wait lists. So that $0 deserves an honest look. Some of it reflected a staff that built what students needed in district. Some of it may have been that a separate placement was close to impossible even when a team thought a student needed one. A zero can mean strong inclusion, or it can mean no options, and the federal data can't tell those two apart. Is the separate-school rate partly a map of where those schools were built?
Is it about how placements are paid for? States fund high-cost placements in very different ways. Does reimbursement for those placements change the decisions a team makes, or the options a district invests in?
Is it about what "80% of the day" actually measures? The federal measure counts time in a room. It can't tell us whether a student is truly included in instruction, has the support to succeed there, or is sitting in the back with a paraprofessional.
What else? Family advocacy, due process patterns, staffing shortages and teacher preparation all likely play a part. I suspect the real answer is several of these at once.
We don't need to answer these today. I am committing to working on them, in the open, in the Observatory, with any of you who want to dig in with me.
Try this with your own district
Open the Special Education Observatory and find your district. No sign-up is needed.
Under "Where students learn," write down your share of students in general education 80% or more of the day and your share in separate day schools, with your state's figures beside them.
Pull your current out-of-district placement list and sort it by the main reason for each placement, such as behavior, medical needs or a specialized program you don't offer.
For the largest group, ask your team two questions: what was tried in district first, and what would it take to build that program ourselves?
Bring those three numbers and your answers to your next leadership meeting. It's one of the most useful conversations a special education team can have, for students and for the budget.
Free tools you can use today
Everything below is free and open on pulseforeducation.com, with no sign-up:
State and district pages in the Observatory, with every figure cited to its source, and a toolkit that turns them into board slides, a one-page brief and a letter to your legislator.
Special education staffing by state, from federal data.
Three calculators: the Provider Workload Calculator, the Vacancy Cost Calculator and the Staffing Request Builder.
Learning: the six-minute lesson The Measurement Gap in Six Minutes, two white papers, and The First 90 Days as a Student Services Director.
The Workload Census, a four-minute anonymous survey for providers. Every response makes the picture of our field's real week clearer.
You can find all of it on our resources page.
What we're adding next
The Observatory will keep growing, and the next additions are the data that may make us uncomfortable. We're working on bringing in suspension and expulsion rates for students with disabilities, restraint and seclusion, and disproportionality by race and disability, from the federal Civil Rights Data Collection and IDEA reports. These are hard numbers to look at. We need to see them clearly to change them.
One more thing
When I became a director, I found out how much I missed having a coach of my own. This November I'm opening a small founding cohort for special education leaders, with live sessions on Tuesday evenings. I'll share the details next week.
In the meantime, I'd love to hear from you. Hit reply and tell me what you think explains your state's place on the map. I read every reply, and the best answers will shape what we study next.
Thank you for reading, and for the work you do for students every day.
~Eric