EXPOSED
The paperwork half of this job — grant narratives, outcome reports to funders, budget spreadsheets, program summaries, policy manuals, intake data analysis — is exactly what current AI drafts at usable quality, and that is a large share of a typical week. What survives is the other half: standing in front of a board or a county funder and owning the numbers, deciding which clients get scarce beds or slots, supervising and disciplining frontline staff, and holding relationships with police, schools, hospitals and donors. Licensure is inconsistent (some roles require an LCSW or clinical credential, most do not), so the regulatory moat is thin and the protection comes from accountability and presence instead.
Headcount grew steadily across the period.
Median pay $67,150 → $80,390 -4.2% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+6.4% 219,800 → 233,900 on the projections basis
Growing, and only partly exposed
The BLS expects +6.4% more of these jobs by 2034, and at 52/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~18,600 openings a year on average, including replacing people who leave.
Case ManagerClub ManagerParole DirectorProgram ManagerWelfare ManagerWelfare DirectorCasework DirectorNonprofit ManagerGroup Home ManagerNonprofit DirectorProgram SupervisorBorough CoordinatorMembership DirectorNon Profit DirectorProgram CoordinatorScout Work DirectorOutreach CoordinatorCase Services ManagerHome Service DirectorServices Case ManagerWelfare AdministratorChild Welfare DirectorField Service DirectorYouth Program Director
Holding it up: judgment & accountability . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier An 11 reflects a week genuinely split down the middle: HUD CoC applications, HMIS outcome tables, Title XX and CDBG reporting, staff schedules and policy-manual updates are all draftable by machine, while the un-automatable residue — walking a case aide through a client death, negotiating a subcontract with a partner agency, testifying at a county commission budget hearing — is real but does not fill the calendar.
Some physical or field component A 9 rather than a 3 because the job is not run from a desk: site visits to shelters and group homes, licensing and fire-marshal inspections, filling in on the floor when a shift is short, driving between satellite sites, and being physically present during a client crisis or a staff incident — but the manager is not the one doing the direct hands-on care.
Certification preferred, not legally required A 5 recognises that most postings ask for an MSW or a bachelor's plus experience with no state licence attached, so nothing personally bars an unlicensed person from the chair; the credential that does exist in some roles — LCSW, licensed administrator for a residential facility — plus mandated-reporter duty under state statute is what lifts it off the floor rather than nothing at all.
Exists to be accountable for ambiguous calls A 14 is earned by the calls no procedure covers: who gets the last bed on a cold night, whether to discharge a non-compliant client, whether an allegation against a staff member goes to CPS and to termination, and which program you cut when a grant is not renewed — each defensible only after the fact, each with a named person answering for it.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (11/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (5/20) is whether the law requires a licensed human to sign. Trust premium (13/20) is whether buyers specifically pay for a person. Judgment and accountability (14/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 32 of this occupation's 52 points (62%).
Embodiment (9/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 68/100 — SAFE.
Funder/insurer terms rather than statute: HUD CoC and SAMHSA grant agreements, or nonprofit D&O and professional liability carriers, requiring a named human program manager to attest that outcome data and eligibility determinations were reviewed by a person, with attestation liability for false claims (FCA exposure already attaches to grant certifications).
Genuine two-tier job: if grant narratives, outcome reporting and budget assembly are absorbed by AI, the residual week is staff discipline, union grievances, board and county-funder confrontation, incident review and interagency negotiation — none of which current systems do. Watch for job postings dropping 'grant writing' from the essential functions while adding supervision and compliance scope.
State licensing of program administrators in regulated service lines — e.g. state child-placing agency and residential facility rules (already in several states' DCFS/DHS licensing codes) requiring a named, licensed program director with an LCSW/LMSW or equivalent who personally signs placement, restraint-review and critical-incident reports. Expansion of that named-responsible-person requirement to homeless CoC-funded, SUD (42 CFR Part 8 opioid treatment program sponsor model) and Medicaid HCBS providers would make a licensed human's signature legally unavoidable on the exact judgment calls AI drafts.
If coordinated-entry and bed/slot rationing decisions become formally auditable — state or CoC policy naming an accountable human for each prioritization override, as some Continuums already require for vulnerability-score deviations — the role's ownership of scarce-resource triage becomes documented rather than informal.
Buyer here is a funder, not a consumer, so the premium is thin. The narrow route: county and state RFPs adding key-personnel clauses that name and price the program manager as a deliverable (common in government contracting), and AI-content disclosure rules in grant applications making human-authored narratives a scored criterion.
The limit. The paperwork share is large and already automatable, so task_resistance has limited headroom. The strongest realistic gains are institutional and uneven — they attach to licensed, regulated service lines (child welfare, residential, OTP) and largely bypass managers at unlicensed community nonprofits, where no plausible lever exists.
| New York-Newark-Jersey City, NY-NJ | 16,090 | $100,570 +25% |
| Los Angeles-Long Beach-Anaheim, CA | 12,400 | $87,100 +8% |
| Boston-Cambridge-Newton, MA-NH | 5,730 | $85,430 +6% |
| Chicago-Naperville-Elgin, IL-IN | 5,390 | $79,150 -2% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 5,320 | $80,810 +1% |
| San Francisco-Oakland-Fremont, CA | 5,080 | $93,410 +16% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 3,570 | $89,990 +12% |
| Dallas-Fort Worth-Arlington, TX | 3,390 | $80,480 +0% |
| Salem, OR | 460 | $110,210 +37% |
| Olympia-Lacey-Tumwater, WA | 220 | $108,380 +35% |
| Seattle-Tacoma-Bellevue, WA | 2,040 | $106,400 +32% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 52. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.