What a Shrinking Behavioral Health Workforce Means for the Members Who Are Already Waiting

A new brief from HRSA's Bureau of Health Workforce puts a number on something health plan leaders have felt for years: the people who deliver behavioral healthcare are running out faster than the people who need it are growing in number. By 2038, the report projects the country could be short somewhere between 99,780 and 203,690 mental health counselors, depending on the scenario modeled. Psychologists are projected to be short by 99,840 to 152,520.

Under the more severe scenarios, the report puts psychologist and mental health counselor supply at well under half of what the population will actually need.¹ That is not a workforce that is stretched thin. That is a workforce that cannot cover the population it is meant to serve.

The wait health plans already manage is about to get longer
Access has been the industry's stated priority for years, and the effort has produced real progress in some places. Unfortunately it has not solved the underlying gap. Today, 40% of the U.S. population, about 137 million people, already lives in a federally designated Mental Health Professional Shortage Area¹, and the national average wait time for a behavioral health appointment is 48 days.¹ Six in ten psychologists are not accepting new patients at all.¹

None of that improves if supply keeps shrinking while demand keeps climbing. HRSA's projections describe a workforce that gets smaller from here, and they assume today's clinicians hold steady, which is generous given that 93% of behavioral health professionals reported burnout in a 2023 survey and 62% reported it as severe.¹ The wait a member faces right now is closer to the best case than the worst one.

Getting a member into care is not the same as getting them better
What is left of the workforce is not evenly distributed, either. In 2021, 69% of rural counties had no psychiatric nurse practitioner at all, compared to 31% of urban counties. For psychologists, that gap was 45% versus 16%.¹ A shrinking supply of generalists does not automatically translate into the right supply of specialists, either. A member with autism, unresolved trauma or a substance use disorder needs a clinician trained to treat that specific condition, not just any available appointment slot.

This is the trap in a shortage environment: pressure to fill capacity fast makes it tempting to measure success by whether someone got an appointment at all. Provider count is not access. Appointment availability to evidence-based care is access. If a scarce workforce gets routed to the wrong care for what is actually driving a member's symptoms, the visit still happened, the spend still hit the ledger and the member is still not better. Everyone gets seen. Few get well.

Unmet needs do not stay unmet. They go somewhere else.
When people cannot get evidence-based care fast enough, they do not always wait. They substitute. More than 40 million people a day already bring health questions to ChatGPT. Among adults with a mental health condition who use AI chatbots, roughly half report using them for therapeutic support, and 36% say they find them more helpful than a human therapist.³ OpenAI's own data shows that in a given week, about 0.15% of active users have conversations containing explicit indicators of suicidal planning or intent.⁴ At that platform's scale, that is more than a million people a week bringing the most acute moment of their lives to a product with no clinical accountability behind it. 

This is not a fringe behavior and it is not a forecast. It is what members are doing right now, while they wait 48 days for an appointment. Nobody knows exactly what that looks like at scale in 2038. What we do know is that every member who defaults to an unproven substitute because the evidence-based option was not easy or fast enough to access is a member whose underlying condition may go unaddressed, driving escalations in symptoms and ultimately cost.

The fix is not convincing members to wait longer for a shrinking workforce, and it is not hoping a chatbot holds the line until enough new clinicians graduate from school. The fix is making the workforce that already exists actually work for the people who need help by improving our methods for fast and effective care matching, so that members can quickly identify the best evidence-based resource for their individual need. Scarcity raises the cost of routing someone to the wrong resource. When capacity is abundant, a mismatch might be a short detour on the member's care journey. But when capacity is this tight, a mismatch costs weeks to reach the wrong provider and weeks again to get back in line for the right one, and not everyone waits through both before turning to chatbots. 

More access to the right care shrinks the shortage. It does not just route around it.
This is the premise NovaOne was built on, and the HRSA numbers make the case sharper. A national shortage of licensed clinicians is not solved by tweaking a traditional directory or adding another point solution competing for the same limited pool of providers. It is solved by getting members to a clinician who treats their actual condition on the first try, so that clinician's limited hours go toward members who get better, instead of members who cycle through care that was never properly matched to what was driving their symptoms in the first place.

Every member routed correctly the first time is capacity handed back to a workforce that HRSA says is about to have less of it. That is not a slogan. It is the mechanism: precise navigation to evidence-based care does not just help the member in front of you, it protects the throughput of a shrinking system for the member behind them. 

No one knows exactly what behavioral healthcare looks like in 2038. But the workforce data for the 12 years between now and then is already written down, and it says the gap widens before it narrows. Health plans that build for evidence-based navigation now, not just access, are the ones positioned to make a shrinking workforce go further, for more members, for longer.

If you want to talk through what this means for your specific population, we would be glad to share what we are seeing. Connect with the NovaOne team here.

Sources:
1.    HRSA Bureau of Health Workforce, Behavioral Health Workforce Brief 2025
2.    OpenAI, AI as a Healthcare Ally, January 2026 
3.    Rousmaniere et al., "Large Language Models as Mental Health Resources," Practice Innovations, 2025 (Sentio University survey, n=499) 
4.    OpenAI, "Strengthening ChatGPT's responses in sensitive conversations," October 27, 2025