In the wake of devastating school shootings, members of Congress from both parties came together and appropriated roughly $1 billion to permanently place 14,000 mental health professionals in the schools that need them most — low-income districts and rural communities.

By the numbers, it worked. In the first year, the programs delivered mental and behavioral health services to nearly 775,000 elementary and secondary students nationwide. Sampled projects reported a 50% reduction in suicide risk at high-need schools, along with drops in absenteeism and behavioral incidents and measurable increases in positive student-staff engagement.

On April 29, 2025, the U.S. Department of Education moved to end it.

On July 29, 2026, a federal judge in Seattle told the Department, for the second time, that it could not.

---

Why this one lands differently in Connecticut

Every state in this lawsuit has an interest in the money. Connecticut has a different kind of interest.

On December 14, 2012, twenty first-graders and six educators were killed at Sandy Hook Elementary School in Newtown. In the years that followed, this state rewrote its gun laws, overhauled school security, and — less noticed at the time — built out mental health provisions in the 2013 legislation because the people who studied what happened concluded that hardened doors alone were not an answer.

The federal programs at issue in this case sit in that same lineage. So the four Connecticut districts standing to lose social workers this fall — Hartford, New Britain, Vernon, and Waterbury — are not an abstraction. They are roughly $3 million in funding through 2029, in a state that has spent thirteen years being told, correctly, that school mental health infrastructure is part of school safety.

The grim pattern behind the funding

The Congressional Research Service published a report last November, R48740, School-Based Mental Health: Introduction and Considerations for Congress. Read it alongside the litigation and one thing becomes impossible to miss: federal investment in school mental health has, almost without exception, arrived immediately after a massacre.

• 1999, Columbine. Congress and the executive branch expand investment through joint Education/HHS efforts, including the Safe Schools/Healthy Students Initiative and Project SERV, which sends funding directly to school communities in the aftermath of a crisis.
• 2012, Sandy Hook. Federal support continues through programs including HHS’s Project AWARE.
• 2018, Parkland. Within weeks of the shooting that killed fourteen students and three staff members, Congress creates the Mental Health Service Professional Demonstration Grant Program (MHSP) with a $10 million appropriation, directing the Department to address the shortage of school-based mental health providers in low-income schools. President Trump convenes a Federal Commission on School Safety, which finds a consistent theme in its listening sessions — limited access to mental health professionals in the highest-poverty districts — and recommends expanding in-school access to care.
• 2020. Congress creates the School-Based Mental Health Services Grant Program (SBMH) at $10 million, to directly hire and retain credentialed providers, with a 25% grantee match so the programs would eventually stand on their own.
• 2022, Uvalde. Nineteen students and two teachers killed at Robb Elementary. Congress responds with the Bipartisan Safer Communities Act, Pub. L. No. 117-159, directing an additional $100 million per year for each program for fiscal years 2022 through 2026.

The Act’s Republican sponsors were explicit about the theory: too often, adolescents with untreated mental health conditions become the people who commit these acts, and the law was written to connect students in crisis with care before it is too late.

CRS also identifies the flaw in the pattern. Federal money for school mental health tends to arrive after catastrophe and recede when attention moves elsewhere — often with no long-term strategy for sustaining what it built. That is the structural problem underneath this lawsuit, and it would exist even if no one had ever tried to cancel a grant.

What the research says about mental health and school violence

If the argument for these programs is school safety, the argument deserves scrutiny rather than assertion. CRS provides it, and the picture is more textured than either side’s talking points.

School-based mental health programming generally follows a multi-tiered system of support. Tier 1 is universal — schoolwide climate policy and social-emotional curricula for everyone. Tier 2 is selective — small-group or brief intervention for students showing risk factors. Tier 3 is indicated — individualized treatment for students with identified needs, usually requiring a licensed clinician. The grants in this lawsuit primarily fund the people who deliver Tiers 2 and 3.

On violence specifically, CRS describes school mental health as operating on both ends: preventing violence, and treating its aftermath. Tier 1 work targets bullying and aggression. Tier 2 includes peer mediation and classroom management for students showing early conflict behavior. Tier 3 addresses students displaying serious aggressive or antisocial behavior.

The most concrete mechanism is the threat assessment model — an empirically supported process in which a trained school team evaluates an explicit threat, determines whether it is transient or substantive, and triages the student toward mental health services, law enforcement, or both. The FBI and Secret Service, working with the Department of Education, have recommended schools adopt this approach. Studies of the Virginia model found students in those schools reported less bullying, a better learning climate, and greater willingness to seek help from adults.

None of that runs without staff. A threat assessment team requires people qualified to sit on it.

On effectiveness generally, CRS reports that well-implemented programs increase prosocial behavior, reduce anxiety and depression, prevent aggression and substance use, and reduce suspensions and disciplinary referrals — with selective and indicated interventions showing larger effects than universal ones.

And the honest caveats, which advocacy documents leave out. Benefits for academic achievement are inconsistent across studies. Simply offering programming guarantees nothing: poorly implemented programs, or programs mismatched to the population served, can be counterproductive and in some cases worsen the outcomes they target. Fidelity to the tested design, adequate training, appropriate staffing, and ongoing evaluation are what separate effective programs from expensive ones. CRS also lays out a serious case against school-based mental health — that emotional development belongs to families and community institutions, that schools face mission creep without matching resources, that teachers are being asked to work outside their training, that services may duplicate what community providers already offer.

Those are not fringe positions. They are arguments about how to do this well, not proof that the shortage is imaginary. On that last point the numbers are not ambiguous: HRSA reports that more than half the U.S. population lives in a mental health professional shortage area, and projects that unmet need for psychologists, school counselors, and school social workers will grow by at least 50% by 2036.

What the Department did

On April 29, 2025, the Department sent “Notice of Non-Continuation of Grant Award” letters to grantees across the plaintiff states. Per the complaint in Washington v. U.S. Department of Education, No. 2:25-cv-01228 (W.D. Wash.), every letter was identical boilerplate. Each said the grant reflected the prior administration’s priorities and conflicted with the current administration’s, then offered a disjunctive list of four possible reasons: the program violates the letter or purpose of federal civil rights law; conflicts with the Department’s policy of prioritizing merit, fairness, and excellence in education; undermines student well-being; or is an inappropriate use of federal funds.

The letters did not say which of the four applied to any particular grant. They did not mention grantee performance at all. Grants were set to terminate December 31, 2025.

The same day, the Department told Congress it was discontinuing approximately $1 billion in awards and planned to “re-envision and re-compete” the money. To the press, per the complaint, the Department identified DEI as the reason — objecting to a competitive preference priority that had awarded bonus points to applicants planning to hire providers from diverse backgrounds or from the communities where they would work.

The legal hinge: continuation awards are not competitions

This is the part most coverage skips, and it is the core of the case.

The Department runs two entirely different processes. For new grants, it publishes an application notice in the Federal Register with that year’s priorities and scoring criteria; applicants tailor proposals; reviewers score and rank. 34 C.F.R. §§ 75.100, 75.101, 75.105, 75.217.

For continuation awards — the annual renewals on a multi-year project already underway — there is no application and no competition. The Department eliminated the continuation application requirement in 1985. Under 34 C.F.R. § 75.253(b), what the Department considers is grantee performance: performance reports, performance measures, financial data. Under § 75.253(c), continuations get priority over new grants. The Department itself described continuation cut-offs as extremely rare in a 1994 Federal Register notice, and said in August 2024 that when they happen, grantees usually see them coming well in advance.

So the question is not whether the administration can set new priorities. It is whether it can apply them retroactively to kill projects already three years into a five-year term, without ever mentioning how those projects performed.

The six counts:

1. APA, arbitrary and capricious (5 U.S.C. § 706(2)(A)) — no individualized assessment (SEC v. Chenery, 332 U.S. 194 (1947): an agency cannot leave a party guessing at its theory); inconsistent application (the Department left grants intact for the Indiana DOE, Fort Wayne Community Schools, and Norman Public Schools that were awarded under the same priorities); and failure to account for reliance interests (FDA v. Wages & White Lion Investments, 145 S. Ct. 898 (2025), quoting Encino Motorcars v. Navarro, 579 U.S. 211 (2016)).
2. APA, contrary to law (§§ 706(2)(A), (C)) — importing new priorities exceeds the Department’s own regulatory authority under § 75.253.
3. Declaratory judgment on what § 75.253(b)–(c) permits.
4. APA, notice and comment (§ 706(2)(D)) — GEPA, 20 U.S.C. §§ 1221e-4, 1232, requires notice-and-comment rulemaking to set grant priorities. The Department followed it in 2022. It did not follow it here.
5. Spending Clause and separation of powers — Pennhurst fair notice, plus the argument that GEPA’s equity directive, 20 U.S.C. § 1228a(b), requires applicants to address barriers to equitable participation. If the Department cut these grants because grantees complied with a congressional mandate, that is a Take Care Clause problem.
6. Equitable ultra vires — Armstrong v. Exceptional Child Center, 575 U.S. 320 (2015).

Count 5 is the one I would watch. Arguing that the executive punished grantees for following a directive Congress wrote is an unusual posture, and it is where the constitutional claims do real work.

The obligation that does not disappear

One point I have not seen made in any of the coverage, and the one most relevant to my practice.

Federal grant money is discretionary. IDEA and Section 504 obligations are not.

Under the Individuals with Disabilities Education Act, psychological services, social work services, and counseling services are “related services” — supports required to help a child with a disability benefit from special education. Under IDEA and Section 504, eligible children are entitled to a free appropriate public education. Where mental health support is necessary for a student to receive a FAPE, the district must provide it, whether or not a federal grant is paying the clinician.

For a district losing grant-funded staff, the questions are:

• Which students on IEPs were receiving counseling or social work services from grant-funded personnel?
• Can the district still deliver those services at the frequency and duration the IEP specifies?
• If not, is it staffing to the IEP — or quietly amending the IEP to match reduced staffing?

That last one is where I expect problems. A district that cuts services because it lost a grant has a funding problem. A district that cuts the IEP to match has a legal problem, and parents have due process rights to contest it. If your child’s services change this fall, get the change in writing and ask what is driving it.

Where the case stands

• June 30, 2025 — Complaint filed in W.D. Wash. Sixteen plaintiff states, including Connecticut (through AAG Andrew Ammirati), Massachusetts, and Nevada.
• November 28, 2025 — CRS publishes R48740.
• December 2025 — The coalition secures an order declaring the discontinuations unlawful, requiring new continuation decisions, and entering a permanent injunction barring implementation “through any means.”
• July 2026 — Rather than making the promised six-month funding determinations, the Department moves to terminate the same grants outright. The coalition files a second suit.
• July 29, 2026 — TRO granted. It expires August 24, 2026.

Three forms of relief are in play and they are not interchangeable. A temporary restraining order is emergency, short-term, and preserves the status quo until a real hearing — this one runs out August 24. A preliminary injunction, which the court is expected to schedule a hearing on, would hold through the litigation. A permanent injunction comes after the merits; the coalition already has one from December, and the current fight is essentially about whether the Department found a way around it.

The TRO does two things: it blocks terminations premised on the new priorities, and it blocks terminations that skip the statutory procedural protections grantees are owed — including the requirement that the agency first work with the grantee to resolve problems. Even where an agency has substantive discretion, process requirements are separately enforceable.

Connecticut reports six favorable rulings against the administration on these grants so far. The 2026 coalition is 15 states; Nevada, a plaintiff in the original complaint, is not on the current list.

What to watch

Circle August 24. If a preliminary injunction issues before the TRO lapses, Connecticut’s four districts get breathing room through the life of the case. If not, terminations go live again and districts start making staffing decisions in the middle of a school year.

For grantees: preserve your continuation correspondence, performance reports, and staffing records. The strongest allegations in this complaint are documentary — identical letters, identical priorities, comparable grants left untouched.

And for everyone else, the observation that will outlast this litigation. The money came after Parkland and after Uvalde. It came after Columbine and after Sandy Hook, under different names. It has always come after. Districts that built five-year staffing plans on it were doing exactly what the federal government told them to do. Whatever happens in August, the deeper problem is that school mental health infrastructure in this country is funded by grief, on a schedule set by tragedy, and that is not a plan.

IF YOU NEED A LAWYER CONTACT US