# Per-State Dartboard Narratives — Youth Mental Health Access Gap V1
**Companion to**: `mh_gap_youth_v1_article.md` §S3
**Source**: `analysis_outputs.mh_gap_youth_v1_dartboard_v1`
**Sampling**: population-weighted, stratified by `residual_class` (target 4 expected + 4 positive + 4 negative; actual 4+3+2=9 due to small outlier classes)
**Each state is pre-specified — population-weighted random sample with fixed RNG seed.**
---
## POSITIVE OUTLIERS — gap worse than uninsured rate predicts
### 1. North Carolina · z = +1.81 · Need: 39.1% · Population under-18: 2.32M
NC is a populous, geographically and economically heterogeneous state with an unexpectedly low youth-serving provider density (9.1 per 100K under-18). The 4.2% uninsured rate would predict a much smaller gap than is observed. The likely structural drivers are (i) the under-investment in publicly-funded community youth mental-health services since the early-2010s reform that closed several state psychiatric hospitals without proportional community supply build-out, and (ii) the geographically uneven distribution of supply (concentrated in the Raleigh-Durham-Chapel Hill triangle and Charlotte metro, with substantial under-supply in coastal and Appalachian counties). State-level analysis cannot separate these geographies, but they are the recurring themes in NC mental-health policy literature.
The replication template that the adult MH Gap V1 paper identified — Medicaid expansion + CCBHC — is *partial* in NC: the state expanded Medicaid in 2023 but has not implemented the CCBHC payment model at scale. The residual flags this gap as larger than uninsured-rate alone explains; CCBHC implementation is the most likely policy lever to close it.
### 2. New Jersey · z = +1.53 · Need: 36.3% · Population under-18: 2.01M
NJ is among the most insurance-saturated states (2.6% uninsured under-19, near the national floor) yet shows an unexpectedly low youth-serving provider density (17.9 per 100K). The need prevalence (36.3%) is on the lower end of the dartboard but the supply shortfall is sharp. The pattern is consistent with **structural under-investment in youth-specific mental-health workforce** independent of insurance status — providers exist in NJ, but the youth-serving subset is thin relative to adult-serving capacity.
Likely policy drivers: New Jersey's mental-health workforce has historically focused on adult populations served by the strong commercial insurance market; pediatric and adolescent specialty supply has lagged. The residual surfaces this for follow-up rather than explaining it.
### 3. Puerto Rico · z = +2.39 · Need: 39.2% · Population under-18: 499K
Puerto Rico is the most extreme positive outlier in the dataset. Youth-serving provider density is 5.8 per 100K under-18 — by far the lowest in the analysis. The 2.5% uninsured rate (the Puerto Rico Medicaid-equivalent ASES program is comprehensive) creates an even larger predicted-vs-observed gap than the raw density alone suggests.
The structural explanation is the post-Hurricane Maria health-system disruption compounded by decades of federal Medicaid-payment cap policies that have constrained Puerto Rico's mental-health workforce build-out. The data does not assign causation but the magnitude (+2.39σ — the strongest signal in the dataset) is consistent with the documented Maria-era and post-Maria workforce migration to the U.S. mainland.
The policy lever for PR is federal: full Medicaid parity for the territory, Title V MCH grant expansion, and IHS-style federal direct-employment of mental-health workforce in underserved territories.
---
## NEGATIVE OUTLIERS — gap better than uninsured rate predicts
### 4. Vermont · z = −1.70 · Need: 29.3% · Population under-18: 115K
Vermont has the *lowest* youth mental-health need prevalence in the dartboard (29.3%, 10 percentage points below the national pop-weighted average) **and** an exceptionally high youth-serving provider density (1,059 per 100K under-18 — among the highest in the dataset). Despite a fine but not extraordinary uninsured rate (3.9%), the gap is 1.7σ smaller than the national regression predicts.
The structural drivers are well-documented: (i) the University of Vermont Medical Center's pediatric behavioral-health program (one of the largest in northern New England), (ii) Vermont's 2014 Medicaid expansion with substantial state-supplementary funding for community mental-health services, and (iii) the state's Designated Agency system — community mental-health centers with statutorily-defined service obligations that include youth-specific service lines.
Vermont is the **policy-driven replication template** for the youth gap: medical school + Medicaid expansion + state-mandated community workforce capacity.
### 5. Alaska · z = −1.69 · Need: 43.2% · Population under-18: 174K
Alaska is structurally different from Vermont and reveals an alternative pattern. Need prevalence is *high* (43.2%) — among the higher in the dartboard, consistent with documented elevated youth distress in rural, isolated, and Indigenous communities. Insurance coverage is below the national average (5.5% uninsured under-19, slightly worse than national).
Despite high need and moderate uninsured rate, the gap is 1.7σ *better* than the regression predicts because **provider density is 1,085 per 100K under-18** — among the highest in the dataset. The driver is the Indian Health Service + Alaska Native tribal health organizations' federally-employed and state-employed mental-health workforce, who are NPI-registered and counted in the supply layer even though they serve populations the state-level supply count rarely sees in other states.
The Alaska case is a federally-built supply system. The replication lever is the Title V MCH block grant + IHS direct-employment model; states with significant Indigenous populations have access to the analogous federal mechanism but it has not been built out at Alaska's intensity elsewhere.
---
## EXPECTED — what the regression predicts holds
### 6. Texas · z = +0.70 · Need: 42.4% · Population under-18: 7.54M
Texas is the largest state in the dartboard by under-18 population (7.5M, ~18% of the entire participating sample). Need is high (42.4%), uninsured is the highest in the dataset (8.9%), and youth-serving provider density is among the lowest of mid-large states (16.9 per 100K). The gap is large — but at z=+0.70 it sits within the expected band, meaning the high uninsured rate predicts most of it.
The policy reading: Texas's youth mental-health gap is not anomalous given its insurance gap. The fix is *both* — Medicaid expansion to close the uninsured gap **and** workforce build-out to close the supply gap. The framework does not flag either as the primary driver; both are structural.
### 7. Utah · z = −0.41 · Need: 37.0% · Population under-18: 934K
Utah has below-average need (37.0%), low uninsured under-19 (4.7%), and average youth-serving supply (180.9 per 100K). At z=−0.41, the gap is slightly better than predicted but well within expected. Utah's case is unremarkable by the residual analysis — the state's combination of high private insurance penetration, integrated faith-community support systems, and moderate state supply produces an outcome consistent with the regression baseline.
### 8. Illinois · z = +0.08 · Need: 38.2% · Population under-18: 2.70M
Illinois sits essentially on the national regression line (z=+0.08). The state has near-universal insurance (2.7% uninsured under-19 — among the lowest), average need (38.2%), and high supply density (145.0 per 100K). The dartboard hit here confirms Illinois as a **baseline expected case**.
### 9. Nevada · z = +0.32 · Need: 44.1% · Population under-18: 685K
Nevada has high need (44.1%, among the higher in the dartboard), moderate uninsured rate (6.6%), and moderate supply (50.1 per 100K). The gap is slightly larger than the regression predicts but well within expected. The Las Vegas-Reno metro concentration of supply, combined with the rural under-supply elsewhere in the state, is invisible at state geography but consistent with the regression's adult-state precedent.
---
## What this dartboard says, taken together
- **Three positive outliers cluster on a single structural pattern**: youth-serving workforce supply that is dramatically under-invested relative to what the insurance landscape predicts (PR is the extreme case; NC and NJ are domestic analogs).
- **Two negative outliers cluster on two distinct effective patterns**: (i) university-anchored small-state Medicaid expansion with mandated community-mental-health capacity (Vermont), and (ii) federally-employed structural supply via IHS + tribal health organizations (Alaska).
- **Four expected cases confirm the framework's ability to identify "on the regression line" places** at state geography. The dartboard's three large expected hits (TX, IL, UT, NV) span the spectrum of insurance + supply combinations and all fall within the expected residual band.
The youth dartboard is smaller than the adult tract-level dartboard (9 hits vs 12) because the negative-outlier class has only 2 states. The framework handles this gracefully by sampling all available states in under-sampled classes rather than producing synthetic noise. The qualitative finding — that the policy stack of Medicaid expansion + workforce capacity + parity enforcement closes the youth gap — is consistent with the adult-level finding and reinforced by the structural-supply pattern in Alaska.