The 2025 data on the direct care workforce is in. The headline most agency operators see in industry coverage — "home care aides are now the largest occupation in America" — is true and not particularly useful. The more operationally useful read of PHI's Key Facts 2025 and the BLS Occupational Outlook Handbook is this: the wage-replacement gap is structural, the projected demand has not narrowed, and the retention math at the median agency hasn't changed enough to cover the gap.
If you run an agency, the numbers below are the basis for any honest conversation about staffing capacity, payroll modeling, and what your retention budget actually needs to be.
The size and shape of the workforce
According to PHI's 2025 Key Facts, the U.S. direct care workforce is 5.4 million workers, distributed roughly:
| Setting | Workers | Share |
|---|---|---|
| Home care workers (HHA + PCA) | 3.2M | 59% |
| Nursing assistants in nursing homes | 492,000 | 9% |
| Workers in residential care settings | 687,000 | 13% |
| Other settings (hospitals, etc.) | 1.0M+ | 19% |
The home care segment alone is now larger than the entire teaching profession in the U.S. and roughly the size of the construction labor force. BLS classifies these workers under SOC 31-1120 (Home Health and Personal Care Aides) and the OOH reports a median annual wage of $34,900 as of May 2024 — but that's the wage figure, not the earnings figure. Median earnings across the direct care workforce, accounting for part-time hours and gaps between assignments, is just under $26,000 per year per PHI 2025.
The gap between the wage figure and the earnings figure is the operationally meaningful number. It tells you that even at a competitive hourly rate, workers in this industry don't get enough hours to clear a livable income. That's a staffing problem disguised as a recruitment problem.
Why most retention programs underperform
PHI's 2025 figures put 36% of direct care workers at or near poverty and 49% on some form of public assistance. That isn't an anomaly — it's the load-bearing reality of the labor pool. When a worker can take a $1.50/hour raise to leave you for a competing agency, two things are true: (1) the raise is real money, but (2) the raise still doesn't put the worker above subsidy cliffs (childcare, SNAP, Medicaid). The retention question for these workers is rarely "can I clear another dollar at the new agency" — it's whether the schedule is more reliable, the supervisor responds faster, and the no-show rate among the agency's other caregivers is low enough that they're not constantly being pulled into double shifts.
The BLS OOH projection is that home health and personal care aide employment grows 17% from 2024 to 2034 — about three times the all-occupations average — with roughly 765,800 openings projected each year over the decade. Combine that with PHI's projection of 9.7 million total job openings in direct care through 2034 (new jobs plus replacements) and the implication for any individual agency is straightforward: the supply side is structurally short and competing agencies will keep poaching anyone you successfully retain.
The retention numbers most agencies actually face
Industry surveys consistently put home care aide turnover near 75% annualized. That number gets used so often it's stopped registering. To make it concrete: an agency with 100 caregivers losing 75% per year is hiring, onboarding, and credentialing 75 new caregivers in a 12-month period to net zero growth. If you assume the median sourcing cost per hire is $1,000–$1,500 (job board fees + recruiter time + onboarding overhead) and that lost productivity during ramp-up is roughly 80 hours per new hire at $20/hour, the annualized churn cost on a 100-aide agency runs $200,000–$300,000 — not counting the revenue lost when shifts go uncovered during the gap.
This is why retention investments usually beat recruiting investments at the margin. A $1.00/hour raise across 100 caregivers with a median 30 hours/week is roughly $156,000/year. If it cuts turnover from 75% to 60% — which is well within the band PHI tracks across regional employer programs — you've avoided 15 replacement hires (~$30K–$45K in sourcing costs) and saved roughly 1,200 hours of lost productivity. The arithmetic favors the raise.
The data we'd want every agency operator to know cold
A short list of numbers worth committing to memory before the next staffing meeting:
- 5.4 million total direct care workers — PHI 2025
- 3.2 million home care workers specifically — PHI 2025
- $34,900 BLS median annual wage, May 2024, SOC 31-1120 — BLS OOH
- ~$26,000 PHI median annual earnings (wages × actual hours worked) — PHI 2025
- 17% projected employment growth 2024–2034 — BLS OOH
- 765,800 projected annual openings through 2034 — BLS OOH
- 9.7 million total direct care openings projected 2024–2034 — PHI 2025
- 49% of direct care workers on some form of public assistance — PHI 2025
- 36% at or near poverty — PHI 2025
Every operational planning conversation — wage scales, the raise budget, the recruiting funnel target, whether to invest in a referral bonus program — should be anchored against these. The data does not improve if you ignore it. The Medicaid 80/20 rule and the CMS Access Rule's Interested Parties Advisory Group requirement are both downstream of CMS having looked at exactly this data and concluded the wage-rate transfer mechanism between Medicaid and direct care workers is broken. Whether or not the 80/20 rule fixes that at the system level, the agency-level math is the same: the labor market this industry is in has not gotten easier and isn't going to.