Every comprehensive evaluation begins the same way. Not with a diagnosis, not with a label, but with a presenting concern — a pattern serious enough, persistent enough, and resistant enough to informal intervention that someone finally puts it in writing and asks for a formal look. This piece is that referral. The subject is not a child. It is the system that 50 years of law, five reauthorizations, and thousands of pages of regulation were supposed to fix.

This is not where a series about special education policy is expected to start. It does not start in 1975, nor does it start with the law. It starts with the question every diagnostician learns to ask before opening a single test protocol: what is actually happening here, and why has nothing tried so far resolved it?

The Presenting Concern

Fifty years after the Education for All Handicapped Children Act guaranteed access to school, the numbers describing what happens once children with disabilities are inside that building have not resolved into a story of success. Nationally, roughly 74% of students served under IDEA exit high school with a regular diploma — a real gain over the exclusion that preceded 1975, and, by disability category, as low as 40% for students with multiple disabilities (National Center for Education Statistics, 2024b). Measured the same way as every other group, the four-year cohort rate was 71% for students with disabilities compared with 87% for all students, and state rates ranged from 84% in Florida to 49% in New Jersey (National Center for Education Statistics, 2024a).

Students served under IDEA make up roughly 14% of public school enrollment. In the most recent national Civil Rights Data Collection, they accounted for 76% of students physically restrained and 68% of students secluded, out of roughly 105,700 students restrained or secluded in a single school year (U.S. Department of Education, Office for Civil Rights, 2025). Forty-five states reported special education teacher shortages in 2025 (Gilmour, 2025). Furthermore, the federal government’s National Longitudinal Transition Study — which follows youth with disabilities into the years after they leave school — was paused for 7 months in 2025, from February through September (Institute of Education Sciences, 2025).

One relevant protection sits entirely outside that visibility. Since 2004, IDEA has barred schools from conditioning a child’s access to school, evaluation, or services on medication (20 U.S.C. § 1412(a)(25)). Restraint and seclusion are tracked nationally through the Civil Rights Data Collection. Even the Office for Civil Rights’ recent complaint dismissals came to light only through a review the Senate education committee’s ranking member requested after the 2025 layoffs (U.S. Government Accountability Office, 2026). No comparable public data collection appears to track how often the medication protection is actually honored. A safeguard Congress deemed serious enough to write into statute has gone for more than 2 decades without the infrastructure to know whether it is being maintained.

None of this is a scandal in the sense of being hidden. It is published, federally collected, publicly available data. A pattern this well-documented, sustained across 5 decades and five reauthorizations, is not a data anomaly waiting for a better year. It is the finding.

What Comprehensive Evaluation Means Here

A comprehensive evaluation does not start by assuming it already knows the answer. It looks at the whole picture — cognitive, academic, behavioral, ecological — because a single data point is never enough to explain a pattern this persistent. It triangulates. It refuses to let one good year, one favorable state, or one success story stand in for the whole record.

This series proposes to run that same evaluation on the special education system itself — not as a rhetorical device, but as a methodology. Later pieces will address the equivalent of a Full Individual Evaluation’s (an FIE’s) cognitive, academic, ecological, and behavioral domains: the philosophical frameworks operating within the system, what it actually produces, the environment it operates within, and what function its most persistent dysfunctions serve. A final piece will do what the closing section of any evaluation report does: state the conclusions the evidence supports and the recommendations that follow, plainly, and stop where an evaluator’s authority stops.

This piece does the work that has to happen before any of that: establishing that a referral is warranted at all. In a real evaluation, that case has to be made on evidence, not urgency. The 50-year record makes it.

Fifty Years of Tiered Intervention

Read chronologically, the history of federal special education law looks like steady improvement: a law passed, then strengthened, then strengthened again. Read the way a diagnostician reads a case file — looking for what was tried, in what order, and with what documented result — it reads differently. It reads like a case cycled through every tier of intervention without ever being referred for the evaluation that tiered system exists to lead to. The following applies that frame deliberately. Congress did not design these reauthorizations as tiers of intervention — but read against the tier logic, they sort themselves with unsettling precision.

The 1975 Education for All Handicapped Children Act was not an intervention in the tiered sense — a targeted response layered onto instruction already happening. It was the universal tier itself. Before it, roughly 1 in 5 children with disabilities in the United States attended school at all (U.S. Department of Education, n.d.). So the law’s remedy was structural: mandate access itself, through the guarantee of a free appropriate public education, due process protections, and the individualized education program (Education for All Handicapped Children Act, 1975). Everything that follows in this history flows from a single fact — access, however necessary, was never going to close the gap on its own.

The 1983 amendments revised and extended the law and made the Office of Special Education Programs the agency responsible for carrying it out (Education of the Handicapped Act Amendments, 1983). The 1986 amendments extended the universal guarantee earlier and wider, requiring states to serve preschoolers with disabilities from age 3 and funding a program that drew states into serving infants and toddlers from birth (Education of the Handicapped Act Amendments, 1986; U.S. Department of Education, n.d.) — the same logic behind universal, front-loaded support in any tiered system: catch it earlier, and less intensive intervention is needed later.

The 1990 reauthorization, which renamed the law IDEA, added autism and traumatic brain injury as eligibility categories and required an individual transition plan for every student moving toward adult life (Education of the Handicapped Act Amendments, 1990; U.S. Department of Education, n.d.). This is targeted, Tier 2 work: specific populations still falling through the cracks of a universal guarantee, and a specific gap — what happens after school ends — that a guarantee of schooling had not addressed.

The 1997 reauthorization intensified that same tier. Congress, drawing on 2 decades of experience, found that identified students too often did not receive instruction in the general curriculum at all (U.S. Commission on Civil Rights, 2002). The fix was more structure: IEPs now had to state how a student would progress in that curriculum, general educators joined the IEP team, and benchmarks and measurable annual goals became required components of the plan (Individuals with Disabilities Education Act Amendments, 1997). This is the recognizable shape of an intervention that is not producing its target outcome — add monitoring, add structure, add accountability — without yet asking whether the model itself needs to change.

By 2004, the system had reached Tier 3 — its most intensive tier. The Individuals with Disabilities Education Improvement Act (2004) aligned special education with the accountability language of the No Child Left Behind Act (2002), required special education teachers to meet a “highly qualified” standard, mandated that services be grounded in peer-reviewed research (20 U.S.C. § 1414(d)(1)(A)(i)(IV)), and — in a detail worth sitting with — barred states from requiring the old discrepancy model for identifying learning disabilities and required them to permit a Response to Intervention process instead (20 U.S.C. § 1414(b)(6); 34 C.F.R. § 300.307(a)). The “highly qualified” definition did not last: the Every Student Succeeds Act removed it from IDEA in 2015 (Every Student Succeeds Act, 2015, § 9214(d)). The irony is almost too precise to need pointing out: the discrepancy model Congress was retiring had already earned a nickname among practitioners — the “wait-to-fail” model — for requiring a child to fall far enough behind before anyone was allowed to act (Lyon et al., 2001). In the same reauthorization where Congress told schools to stop waiting for children to fail before intervening, the federal system had been running its own wait-to-fail cycle for 3 decades and counting.

And then the cycle stopped, without ever completing itself. IDEA has not been reauthorized since 2004 (Dragoo, 2026). Twenty-plus years is a long time to run a Tier 3 intervention without checking whether it is working — and by 2026, the record showed that largely it was not: only 20 states earned the Department’s own “meets requirements” determination that year (U.S. Department of Education, Office of Special Education Programs, 2026a), 45 states had reported special education teacher shortages the year before (Gilmour, 2025), and the presenting concerns this piece opened with had not resolved. That determination figure carries a footnote worth holding on to: six of the matrix’s results elements are scored by ranking the states into thirds, so a third of the states earn a zero on each of them no matter how their students actually did, and a state in the bottom third across the board cannot reach the 80% the determination requires even with perfect compliance (U.S. Department of Education, Office of Special Education Programs, 2026b). The Advocacy Institute (2026) goes further, calling it mathematically impossible for every state to meet requirements in the same year. The instrument built to measure the system’s compliance was designed so someone must always finish last — a detail this series will return to when it asks what function the system’s design serves.

In a properly run intervention process, this is the exact point at which a case gets referred for comprehensive evaluation. Twenty years of a Tier 3 intervention that is not closing the gap is not a reason to try the same tier again. It is the definition of a referral trigger. That referral has not happened.

What has happened instead, as of 2026, looks less like escalation and more like discontinuation: a federal Office for Civil Rights that received more than 9,000 discrimination complaints between March and September of 2025 and dismissed roughly 90% of the more than 7,000 it resolved (U.S. Government Accountability Office, 2026), alongside proposals to loosen the funding itself: the President’s fiscal year 2026 budget asked Congress to fold the IDEA programs into a single “simplified funding program,” and Project 2025 calls for converting most IDEA money into a “no-strings” formula grant (Kolbe & Dhuey, 2025).

A system that has spent 50 years intensifying its own intervention, one reauthorization at a time, does not arrive at “stop monitoring entirely” as the next logical tier. That is not Tier 4 — that is a case being closed without ever having been properly evaluated.

What the System Absorbs

These reauthorizations consistently produced real change in statute while leaving so much of a child’s daily school experience untouched. It comes into focus from inside an ARD room rather than from reading the amendments alone — because absorption happens where implementation lives, at the state level, where federal and state reform alike arrive as recoding. Texas, where I practice, is supplying three live examples. Dyslexia instruction moved from a Section 504 pathway into special education for students who need it as specially designed instruction (Tex. Educ. Code Ann. § 29.0031, West 2026) — and that move deserves a moment of candor. Dyslexia intervention was always specialized instruction; classifying it under 504 simply let states keep it off the special education books. The reclassification changed real things — evaluation rights, ARD committee oversight, due process protections — and those procedural gains matter. What it did not change was the instruction. The same interventionists deliver the same structured literacy programs to the same children; what moved was the paperwork pathway and the honesty of the count. Instructional Arrangement is becoming Intensity Level — in the 2026–27 transition year, every student receiving special education services in Texas is recoded from the old system into the new one while the funding transition completes (Tex. Educ. Code Ann. §§ 48.102, 48.1022, West 2026; Texas Education Agency, 2026). Non-Categorical Early Childhood has become Developmental Delay — a genuinely wider category on paper, reaching to age 9 instead of 5 (19 Tex. Admin. Code § 89.1040(c)(13), 2026) — while children who carry the old label keep it until their next re-evaluation, and the services attached to either label look identical from the child’s side of the table. In each case, the change is real in statute and in records, and largely invisible in what a struggling child receives on Monday morning. For the adults implementing it, each change is disruptive — new forms, new training, new compliance deadlines. For the child, nothing moves. The system does not resist reform. It absorbs it, renames it, and continues operating exactly as before.

The clearest example may be the one Congress offered as the cure. Response to Intervention was enacted into statute in 2004 as a replacement for wait-to-fail — a way to act sooner. Within 7 years, the Department of Education had to issue a formal memorandum reminding districts that an RTI process cannot be used to delay or deny an evaluation (U.S. Department of Education, Office of Special Education Programs, 2011) — a memo that exists only because the practice it prohibits had become common. The reform designed to speed identification had been absorbed into a new, better-documented way to defer it.

That absorption is not incidental to the tiered-intervention pattern above — it may be the mechanism behind it. A system built to convert every reform into documentation rather than into changed practice will keep producing the same outcomes no matter how many times the reform layer is upgraded. Later pieces in this series will name that function directly and ask what it is protecting the system from. For now, the observation stands on its own: reform absorption is itself evidence for why evaluation, rather than another reform, is the right next step. Reform has been tried. Repeatedly. Documented. It has not closed the gap it targeted, and there is reason to suspect it structurally cannot.

What Standard This Evaluation Is Held To

A comprehensive evaluation does not compare a child to an invented ideal. It compares them to a standard — grade-level curriculum, same-age peers, the child’s own prior performance — because convergence of evidence only means something once one knows what the evidence is converging toward. The same requirement applies here. This series measures the system against three standards at once: the purposes Congress itself wrote into the statute at 20 U.S.C. § 1400(d), the legal floor the Supreme Court articulated in Endrew F. v. Douglas County School District (2017), and the record of other nations that faced the same design question — access versus outcome, medical versus educational — and chose differently, with decades of results to show for it. A fourth standard runs underneath it all: decades of direct field evidence from inside the rooms where these decisions are actually made, which either confirms what the statistics show or complicates them in ways the paper trail alone cannot.

Four domains of analysis are coming in the pieces that follow this one — the philosophical frameworks the system operates on, what it actually produces, the environment it operates within, and what function its dysfunction serves — and a closing piece that states the conclusions and recommendations they support. None of those arguments stand alone. Together, measured against the standards above, they form the comprehensive evaluation this piece requests.

The Referral Is Filed

A referral does not conclude anything. It opens a case. What follows in this series is the evaluation itself — domain by domain, as any comprehensive evaluation proceeds — until there is enough convergent evidence to state conclusions about where the system actually is, and enough evidence after that to make recommendations worth calling ambitious.

Fifty years, five reauthorizations, and a well-documented, uncontested, entirely public record of persistent gaps is the presenting concern. It has never been formally evaluated as a whole. This piece is the paperwork that starts that process.

Related Hub Tools

Eligibility Suite · Child Find Reference · SLD Approaches Reference · Texas FIIE Overview Chart