The Handoffs Nobody Owns

Anyone may refer a child to Early Start — including the parent, with no doctor and no diagnosis. The regional centre must complete assessment and hold the first plan meeting within 45 calendar days. Evaluation, assessment and service coordination are free, by federal requirement, at every income level.

Caregivers notice a difference at 1.75 years, raise it with someone at 2.50, and reach a diagnosis at 4.20 (Martin et al., Frontiers in Psychiatry 2026). Latino children are diagnosed about 2.5 years later than white non-Latino children (Zuckerman, Academic Pediatrics 2014).

Nothing in that gap is a violation. Screening at 18 and 24 months is a professional recommendation, not a duty with a consequence, and no party is assigned the job of telling a worried parent that the door is already open to them.

Early Start Family Resource Centers — free, staffed by parents of children with disabilities, and open to a family whether or not the child is eligible for anything. That last clause is the most useful unknown fact in the system.

But they are not an entitlement. Gov. Code §95024 says the lead agency may allocate funds to family resource services out of the federal Part C money remaining after the mandatory components are paid. High priority, discretionary residue. Nothing requires one per catchment.

California is stricter than the federal floor. 17 CCR §52112 requires the regional centre to notify the school district and hold the transition meeting before the child is two years and nine months, with the referral to the district sent by that same date. Federal 34 CFR §303.209 requires the conference not fewer than 90 days before the third birthday.

At three, the eligibility test changes from a 25% delay in one area to significant limitations in two major life activities (provisional, ages 3–5) to substantial limitations in three or more for full Lanterman eligibility.

And the payer flips. W&I §4659(d)(1): a regional centre “shall not purchase medical or dental services for a consumer three years of age or older unless the regional center is provided with documentation of a Medi-Cal, private insurance, or a health care service plan denial.”

As of February 2025 California had 6,929 Early Start children still on the books past their third birthday — a 22.5% increase in one year (DDS Directive D-2025-Early Start-001).

The US Department of Education rated California’s Part C programme “Needs Intervention” in June 2025, with a score of 35.42% — and zero on both transition indicators, because the data were unreliable. The state cannot currently measure its own handoff.

In our own reading of the vendor lists across the regional centres that publish them, 628 organisations are reachable for birth-to-three therapy and 465 at three and over — but only 4 appear in both. Supply falls 26%; continuity falls to nothing. A family will almost certainly change providers at three, not because supply vanished but because the two workforces are different organisations doing different work under different billing codes.

The service coordinator, on paper. The Family Resource Center does transition support — and then its own mandate ends, because the FRC serves families of children under 36 months. The one body that has been helping you is aged out on the same day your child is.

W&I §4647(b): “The regional center shall assign a service coordinator who shall be responsible for implementing, overseeing, and monitoring each individual program plan.” This is the delivery guarantee, and it exists.

24.7% of regional centre clients aged 3–21 received $0 in purchased services in FY2024–25 — 66,618 of 269,992 people. For ages birth to two the same figure is 3.2%.

Statewide, only 67% of authorised service dollars were actually spent; for ages 3–21 it was 56.6% (DDS Annual Purchase of Service Report, FY2024-25). That 43-point gap is the measurable footprint of services that exist in a plan and never reach a child.

A plan listing a service the child never received is not an edge case. It is a quarter of the school-age caseload.

Nobody, in practice. The person statutorily assigned to monitor implementation is carrying sixty-two children, works for the agency whose purchasing decision is frequently the reason the service did not arrive, and the consequence for a regional centre that breaches the ratio for two consecutive reporting periods is a plan of correction.

The health plan must cover behavioural health treatment for autism (SB 946 → Health & Safety Code §1374.73) — but the mandate binds plan contracts and insurance policies, and a self-funded employer plan is neither. Federal ERISA preemption puts it out of California’s reach entirely (29 U.S.C. §1144). In California, 41.7% of private-sector enrollees at establishments that offer coverage are in self-insured plans — 58.2% at firms of 1,000 or more (AHRQ MEPS-IC 2020).

The regional centre may not purchase a service available from a health plan where the family “chooses not to pursue that coverage” (W&I §4659(c)). It may pay a copay or deductible — but only under 400% of the federal poverty level (§4659.1).

The school district owes only what the child needs to access their education.

Each party can point at another, correctly, under its own statute. There is no arbiter with jurisdiction over all three. The family is the integration layer.

Where the dispute is with a health plan there is a real remedy, and it works: of 5,204 Independent Medical Reviews closed in 2024, 51% were overturned by the reviewers and a further 22% were reversed by the plan before the review even happened — about 73% of families got the service (DMHC 2024 Annual Report). It is free.

But a family must know which regulator covers their plan, and DMHC publishes no way to find out. You discover you filed with the wrong agency by filing with the wrong agency.

DMHC’s Help Center, for the plans it regulates — free, and it recovered $9.3m for members in 2024. It has no jurisdiction over the regional centre or the district, and none at all over a self-funded plan.

Two deadlines, and they are not the same one. 60 days from the Notice of Action to appeal at all. 30 days, and before the action takes effect, to keep the services running while you appeal (DDS, Aid Paid Pending fact sheet). A family that files on day 35 keeps its appeal and loses its services in the meantime.

A rights complaint under W&I §4731 goes to the director of the regional centre being complained about, who has 20 working days to propose a resolution — and if the family does not escalate, that proposed resolution takes effect on the 20th working day by default.

Appeals work. Of 427 eligibility issues raised, 141 families were found eligible and 86 got the assessment they had been refused, against 108 found ineligible — roughly 53% ended with the family getting what had been denied (DDS Appeal Requests and Resolutions Report, FY2021-22).

Per-capita appeal rates vary 19-fold between regional centres — from 0.74 to 13.91 per thousand clients. That is not a difference in who has cause to appeal.

Rights complaints rose 185% from FY2020-21 to FY2023-24. Hispanic families are 43.9% of the caseload and file 34% of complaints.

And DDS has published no appeal outcomes since FY2021-22, so the success rate that would persuade a family to file is four years old and getting older.

OCRA — free, one clients’ rights advocate per regional centre, and independent of it by statute. But OCRA states plainly that it weighs the merits, whether you could self-advocate, and resource availability before taking a case: “CRAs can only accept a limited caseload.” Most families get advice, not representation. And OCRA is funded by DDS, the department that oversees the centres it argues against.

Nothing, in the sense that matters. The educational entitlement ends. Adult services are discretionary and rationed. No party inherits the child.

26% of autistic young adults received no services at all in their early twenties. 58% had ever worked for pay after high school — against 92% for learning disability and 84% for speech-language impairment. 37% were “disconnected”: never worked, never continued education, against 7–8% for those other groups (Drexel National Autism Indicators Report, 2015).

The share of families reporting that obtaining services required significant effort rises from 50% at age 17 to over 70% after high school. About 80% of California’s autism caseload is under 22 — the cliff is the shape of the system, not an event at its edge.

The regional centre remains, for those who hold full Lanterman eligibility. Per-consumer spending at 22+ is where the equity gap is widest in the entire system: $70,269 for white consumers against $39,824 for Hispanic consumers.