Belonging and student mental health: what campus programs do

Two findings set the boundaries of this conversation. Nationally, 50% of college students rate their mental health as fair or poor (a Steve Fund survey). In our State of Higher Ed 2026 research, 35% keep their mental-health struggles to themselves.

The second number is the operationally important one. A student who tells no one is invisible to every system an institution has built, all of which are triggered by disclosure. That is the specific gap a belonging structure can address, and it is a narrower claim than the one this topic usually attracts.

Start with what a co-curricular program is not

This needs stating plainly, because the alternative is a program that overpromises and a staff that gets asked to do clinical work it is not licensed to do.

• It is not treatment. A peer group is not therapy, a facilitator is not a counsellor, and no amount of structure substitutes for clinical care.
• It is not a screening instrument. Students in a program are not assessed, and nothing about participation constitutes a risk evaluation.
• It does not resolve the underlying pressures. 59% of college students nationally have considered dropping out because of financial stress (Ellucian, 2024). Belonging does not pay tuition.
• It is not a substitute for capacity. If counselling has a four-week wait, engagement programming does not shorten it.

Any program presented as addressing student mental health should be able to state those four limits without hedging. If it cannot, that is a reason to be careful with it.

What it does change: who notices, and how fast

The realistic contribution is visibility, and it operates through three concrete mechanisms.

A small group notices absence. Six students with a standing commitment and a shared goal register a missing member faster than any early-alert dashboard, because attendance is not the signal: expectation is. Somebody was expecting that person on Wednesday.

Peers reach a withdrawn student more credibly than an office does. A message from a group member lands differently than an email from a dean, particularly for a student whose reason for not disclosing is that they do not want to be a case.

Structure makes referral routine rather than exceptional. The point of a trained facilitator is not to counsel. It is to know where the door is and to walk someone to it without making it an event.

Students are unusually receptive to institutional effort here: 92% say they feel their professors and administrators care about their career success, and 36% say they need to spend more time improving their mental health. The goodwill and the self-awareness are both present. What is frequently missing is a structure that connects them.

Belonging is a design output, not a mood

Students name belonging directly when asked. In the same research, when asked what they wish their school would provide, 27% said a greater sense of belonging, behind more career-focused courses at 32% and just ahead of more career resources at 26%. Worth noting alongside it: 45% said they already receive everything they need, so the students asking are a specific minority rather than the whole population.

Belonging is produced by structural features, not by messaging. Four are worth designing for deliberately.

• A fixed group. Rotating membership destroys the mechanism, because absence is only legible among people who expect you specifically.
• A role. A student with a responsibility has a reason to return that survives a bad week. Passive attendance does not.
• A purpose beyond belonging itself. Groups organized around connection dissolve; groups organized around a task produce connection as a byproduct. Goal-setting and project structures work for this reason.
• A named adult. One staff member the student can identify by name. Not a department, a person.

Where the care has to go

Three practical safeguards belong in any engagement program that operates near this territory.

Train facilitators on referral, not intervention. The whole curriculum is: recognize that something is wrong, do not diagnose it, know the three places to route it, and tell your advisor. Student leaders should be explicitly told that assessing a peer is not their job.

Write the escalation path down before the term starts. Who is called, in what order, and what is documented. A path invented during an incident is not a path.

Do not make wellbeing a program metric. Reporting mental-health outcomes from a co-curricular program invites a claim you cannot support and creates pressure on student leaders to produce it. Measure reach, depth, and repeat engagement. Leave clinical outcomes to the clinicians who can measure them.

The capacity context

Administrators see this clearly and are not resourced for it. 100% say their students worry about financial stability, only 57.9% say their school adequately prepares students for it, and just 32.9% say their budget is adequate to serve student needs.

That is the honest case for peer structure. Not that it is better than professional support, but that at fixed headcount it produces a layer of noticing that staff cannot produce at scale, and it routes students toward professional support faster than waiting for disclosure does. Student-led models exist for this reason: the NSLS chapter model, for instance, is student-led with a published advisor commitment of 10 to 20 hours per semester, which is what makes a noticing layer affordable at all.

Do this next

Write down, on one page, the referral path a student facilitator should follow when a peer stops showing up: who they tell, within what timeframe, and what gets recorded. Then confirm every current facilitator has read it.

That page is the difference between a belonging structure that helps and one that quietly holds a student who needed something it was never able to give. For the wider design principles (the populations conventional programming misses, and how to measure engagement without counting attendance), see our guide to student engagement and belonging in higher education.