This piece quantifies the global mental health treatment gap using WHO Mental Health Atlas 2024 data and asks whether task-shifting to lay health workers -- the field's favorite scale-up strategy -- is actually structured to close a spending gap this large, or just to make the gap look smaller on a dashboard.

I pulled per-capita mental health spending and workforce density figures from the WHO Mental Health Atlas 2024 (144 countries) and the WHO's September 2025 global service-availability release, cross-checked against Our World in Data's synthesis of the same Atlas series.

For the proposed fix, I reviewed the Friendship Bench cluster-randomized trial in Harare (N=576, published via PMC) and a 2024 systematic review/meta-analysis of transdiagnostic interventions delivered by non-specialist providers across 20 LMIC studies.

I did not run new statistical analysis; this is a synthesis and gap-audit of existing peer-reviewed and UN-agency data, so treat the arithmetic as illustrative rather than a formal cost-effectiveness model.

Here is the number that should be the only number anyone quotes in a mental health policy meeting: high-income countries spend $65 per person per year on mental health. 04.

That is not a rounding difference, that is three orders of magnitude, and mental health has sat at 2% of total health budgets since WHO started tracking this in 2017 -- unchanged, un-moved, immune to a decade of awareness campaigns.

I ran this by Isabel Vargas at COLMEX, who does comparative social policy for a living, and her first response was 'you're describing a budget line that doesn't exist, not a budget line that's underfunded.' 04 per capita isn't a policy failure, it's the absence of a policy.

The workforce numbers tell the same story from a different angle. Global median mental health workforce density is 13 per 100,000 people. 1 per 100,000 in much of sub-Saharan Africa -- against roughly 70x that in high-income countries. 3 nurses per 100,000.

A Lancet Psychiatry estimate cited in the Atlas puts 'minimally adequate treatment' for major depressive disorder at 9% globally, 27% in high-income countries, and 2% in Sub-Saharan Africa.

So the treatment gap isn't a gap, it's a cliff, and the cliff correlates almost perfectly with GDP per capita, which should embarrass anyone who still talks about mental health as a stigma problem first and a resource-allocation problem second.

Task-shifting -- training lay health workers to deliver structured psychotherapy -- is the field's answer, and it's a genuinely good answer as far as it goes. 5 in the intervention arm. A 2024 meta-analysis across 20 transdiagnostic studies in LMICs found the approach reliably effective for common mental disorders. I am not disputing the effect size.

I am disputing the implied conclusion that task-shifting is a substitute for the missing budget line, rather than a workaround for it.

Here's the part the scale-up literature keeps burying: the same review base that validates task-shifting also flags misdiagnosis risk, provider burnout, and a near-total absence of studies following outcomes past two years.

Nobody has published a rigorous answer to what happens when a lay health worker delivering PST for 40 clients starts showing the same burnout symptoms she's treating -- because almost no one has funded anyone to look. Task-shifting was designed as a stopgap for physician shortages in the 1980s HIV response. 04 per capita.

Valentina Cruz at UNAM made the ecological comparison I wish I'd thought of first: this is triage dressed up as a treatment model, and triage protocols that run for forty years stop being triage. They become the standard of care by default, and defaults are exactly what stop getting re-examined.