# The Youth Mental Health Access Gap Is Two and a Half Times the Adult Gap — and the State You Live In Is the Binding Constraint
*A long-read working-paper summary from Trellison Institute. May 2026.*
---
The Centers for Disease Control track a quiet number in their biennial high-school surveillance survey. It is the share of American high-school students who, in the past 12 months, experienced two weeks or more of persistent sadness or hopelessness — long enough that they stopped doing some usual activities. The most recent national pop-weighted figure is **39.4%**. Roughly 16.5 million young people across the 35 states whose data was released for 2023.
Thirty-nine point four percent. Approximately **two and a half times the adult rate** of frequent mental distress documented in our companion paper from earlier this month.
We mapped this against the federal registry of licensed mental-health providers filtered to those trained to serve children and adolescents. We asked: where do they live, where can a young person actually get an appointment, and what does the data say about who has built the workforce to meet the need?
The answer is structurally different from the adult case, and the policy implications are sharper.
---
## The 85× problem
Among the 35 participating states, the density of youth-serving mental-health providers per 100,000 under-18 ranges roughly **85-fold**:
- **Alaska** has 1,085 providers per 100,000 under-18 — among the highest in the dataset.
- **Vermont** has 1,059.
- **Delaware** has 1,049.
- **Hawaii** has 983.
- **Maine** has 884.
Then:
- **Texas** has 16.9.
- **New Jersey** has 17.9.
- **North Carolina** has 9.1.
- **Puerto Rico** has 5.8.
This range is approximately three times wider than the comparable adult-serving provider range. The youth-serving mental-health workforce is dramatically more uneven across states than the adult-serving one — because the youth-serving workforce is more *policy-built* than market-driven. Where the policy stack is in place, supply saturates. Where it isn't, supply lags structurally.
---
## What's working
Our analysis identifies two states where the gap is unexpectedly *small* given their insurance landscape — what we call negative outliers. Both are policy stories, not market stories.
**Vermont** has the lowest youth mental-health distress prevalence in the data (29.3%, ten percentage points below the national pop-weighted average) and an exceptionally high youth-serving provider density. The reasons are well-documented: the University of Vermont Medical Center's pediatric behavioral-health program is among the largest in northern New England; Vermont's 2014 Medicaid expansion included substantial state-supplementary funding for community mental-health services; and the state's Designated Agency system — community mental-health centers with statutorily-defined service obligations — includes youth-specific service lines as a matter of state law, not market preference.
**Alaska** is structurally different. Need is *high* there (43.2%, among the higher in the data — reflecting documented elevated youth distress in rural, isolated, and Indigenous communities). Yet the gap is dramatically smaller than the SES proxy predicts because **the youth-serving provider density is 1,085 per 100,000 under-18** — driven by the Indian Health Service plus Alaska Native tribal health organizations whose federally-employed and state-employed mental-health workforce is registered with NPIs that our supply layer counts. This is structural federal supply at work. In states without significant Indigenous populations or without the Title V MCH block-grant footprint that Alaska leverages, the comparable structural-federal-supply pattern is rarer or absent.
Both Vermont's and Alaska's patterns are *replicable*. The federal levers — BHWET workforce expansion, CCBHC payment model with explicit youth-serving certification, Title V MCH block grants — are available to every state. The choice to build the workforce is policy-driven, not demographics-driven.
---
## What's not working
Three states stand out as *positive outliers* — places where the gap is unexpectedly worse than the insurance landscape predicts.
**Puerto Rico** is the extreme case. The youth-serving provider density is 5.8 per 100,000 under-18 — by far the lowest in the analysis. The 2.5% uninsured rate (Puerto Rico's ASES Medicaid-equivalent program is comprehensive) creates an even larger predicted-vs-observed gap than the raw density alone suggests. The structural driver is the post-Hurricane Maria health-system disruption in 2017 combined with decades of federal Medicaid-payment cap policy that has constrained Puerto Rico's mental-health workforce build-out. The data does not assign causation but the +2.39σ signal — the strongest in the dataset — is consistent with documented Maria-era workforce migration to the U.S. mainland.
**North Carolina and New Jersey** are more puzzling. Both have strong adult mental-health workforces. Both have low uninsured rates. Yet the youth-serving subset of their workforce is thin relative to what the adult-serving capacity would predict. The most likely explanation is that the youth-serving mental-health workforce has historically been a separate workforce line from the adult-serving one — different residencies, different practice patterns, different reimbursement dynamics — and that in states where adult-serving workforce build-out has been the priority, the youth-serving lag persists even where adult capacity is strong. Both NC and NJ would benefit from explicit youth-serving CCBHC certification and BHWET workforce expansion targeted at the under-18 panel.
---
## The framework, not the study
We are publishing this as a *working paper* from Trellison Institute, the methodology-audit arm of the DaedArch platform. The same two audit gates that govern the adult companion paper apply here: Trellison's methodology rating of its own analysis, and LedgerWell Corporation's evidence-chain certificate that the audit was carried out faithfully on every analytical step.
Why bother? Because **the methodology is the product**. The Need-vs-Access Framework v1 — which produced the adult tract-level analysis last month and now the youth state-level analysis — is a parameterized reusable tool. The single line change to enable the state-level study was a new `regression_grouping="national"` parameter, replacing the within-state OLS that worked at tract resolution with a single national OLS that works at state resolution. No other code changed.
Eleven more access-domain studies are queued for the same framework: poverty safety-net, English-language acquisition, jobs vs job seekers, postsecondary access, library access, police per capita, maternal care, dental care, broadband, oncology, and crisis response. Each will publish the same content arsenal — working paper, methodology supplement, dataset CSV, data dictionary, dartboard narratives, executive brief, press release, slides, replication package, animated visualizations — under the same audit posture. The first two are now public.
---
## Two policy roads
For states that have not built youth-serving workforce capacity — Puerto Rico above all, but also North Carolina, New Jersey, Texas, and most southern states — the policy work is clear:
1. **Workforce build-out**: BHWET expansion with youth-specific tracks; state-level loan-repayment programs targeting community youth mental-health workforce; Title V MCH block grant youth mental-health line.
2. **Payment-model reform**: CCBHC certification with explicit youth-serving capacity criteria; Medicaid expansion in the non-expansion states; MHPAEA parity-law enforcement on commercial youth benefits.
3. **For Puerto Rico specifically**: full federal Medicaid parity for the territory; an IHS-style federal direct-employment model for mental-health workforce in underserved territories; Title V MCH grant expansion.
For states modeling Vermont's success: state-mandated community-mental-health capacity (Designated Agency or equivalent statute) with Medicaid expansion as the financing layer.
For states modeling Alaska's success: federal IHS + tribal health organization infrastructure leveraged where relevant for any state with significant Indigenous population.
---
## What the data does not say
We do not claim the framework's residual analysis is causal. The negative-outlier states share documented policy stacks, but the residuals only show correlation between policy choice and reduced gap, not that one caused the other. Replication is the test: a state that adopts Vermont's policy stack and shows the gap close over time would be the causal evidence. We provide the data, the framework, and the analytical infrastructure to track that.
We also do not claim youth mental-health is *only* a policy problem. The underlying drivers — social media, academic pressure, post-pandemic isolation, climate anxiety, structural racism, family economic precarity — are the substrate. What the data does say is that *given* the substrate, policy choices determine whether the workforce that responds to it is adequate to the need. Vermont's choice has been one answer; Puerto Rico's circumstance has been another. The 85× supply range is the indictment.
We published the working paper, the methodology supplement, the per-state codebook, the nine case-study narratives, the executive brief, the press release, the replication package, and the slide deck as a single content arsenal at:
**https://trellison.com/research/youth-mental-health-supply-demand-gap**
The full dataset — 35 states × 12 fields, sha256-stamped, CC-BY-4.0 — is downloadable from the same page.
The framework will release V1.1 of the youth paper once the Trellison rating and the LedgerWell certificate are issued. It will then run against its third access domain.
The audit is the product. The framework is the proof. The youth case, taken together with the adult case from last month, says we know what to do — and the methodology says it's reproducible.
---
*Trellison Institute · methodology-rated data journalism · trellison.com*
*Newsletter version (~1,400 words) derived from the v1.0 working paper. The peer-review article is ~3,800 words.*