In 2022, the answer to runaway labor spend was simple: cut travel nurse rates. Facilities did it, rates fell roughly 40% from their pandemic peak, and a lot of CFOs declared victory.
Then the line item kept growing.
If you’re still modeling the total cost of workforce in healthcare the way you did three years ago, you’re probably tracking the wrong variable. Crisis rates were a spike. What replaced them is structural: internal float pools with real overhead, per-diem pools that behave like contingent labor but sit in your core budget, benefits costs rising 7-9% a year, and overtime that quietly backfilled the agency hours you proudly eliminated.
So the spend didn’t leave. It moved.
That’s the shift most employers and facilities haven’t adjusted for. Your dashboard says agency usage is down. Your cost per worked hour says something else. This post covers what changed between 2023 and now, which pieces of conventional advice have gone stale, and what to measure instead. No theory. Just the specific line items that moved and where to look for them.
Three Things That Changed Since 2023
Contract labor stopped being the whole story. For most acute care facilities, agency spend now sits somewhere around 4-6% of total labor cost, down from the 11-15% range at peak. That’s real progress. But the facilities that cut hardest saw overtime jump 20-30% in the same period. You can’t celebrate one without reporting the other.
Benefits became the fastest-growing component. Employer health plan costs rose about 9% in 2025, the sharpest jump in over a decade. GLP-1 coverage, specialty pharmacy, and higher utilization are all pushing it. If your fully loaded cost multiplier is still the 1.25x you set in 2019, it’s wrong by a meaningful margin. Most facilities are closer to 1.33-1.40x now.
Internal pools got expensive. Everyone built one. Few costed one properly. A float pool carries recruiting spend, premium differentials, scheduling labor, and idle time when census dips. Compare it against agency honestly and the gap is narrower than the pitch deck suggested.
The advice that’s now outdated
- “Get agency to zero.” Zero is usually more expensive than 3-5%. You end up paying overtime premiums and eating turnover costs instead.
- “Benchmark on bill rate.” Bill rate tells you almost nothing about delivered cost. Fill rate, time-to-start, and cancellation behavior matter more.
- “Track labor cost monthly.” Monthly closes are too slow for weekly scheduling decisions. You’re reacting a month late.
- “Hours per patient day is enough.” HPPD ignores who worked the hours and at what premium.
Where the Money Actually Sits Now
Here’s a rough allocation for a mid-sized acute facility, based on how these components typically break out:
| Component | Share of total labor cost | Direction since 2023 | The catch |
|---|---|---|---|
| Core staff wages | 58-64% | Up 4-5%/yr | Market adjustments are now annual, not every 3 years |
| Benefits & payroll tax | 22-28% | Up 7-9%/yr | Fastest-rising piece, often modeled with stale multipliers |
| Overtime & premiums | 6-11% | Up sharply | Absorbed the agency reduction; frequently uncounted |
| Agency & contract | 4-6% | Down ~50% | Still the only line most dashboards watch closely |
| Turnover & vacancy cost | 3-7% | Flat to up | Rarely booked to labor at all |
Look at that table and the point gets obvious. The two columns growing fastest are the two most facilities measure worst.
Why Your Current Reporting Misses It
Most cost-of-workforce reporting was built to answer a 2021 question: how much are we spending on travelers? It pulls agency invoices, compares them to budget, and stops.
The problem is that nothing in that pipeline sees a nurse picking up a sixth shift at time-and-a-half, or a per-diem working near-full-time hours without benefits loading, or a vacancy costing you three months of overtime before it’s filled. Those live in scheduling systems, payroll, and the ATS. Three systems that usually don’t talk.
What to do instead: build your total cost of workforce in healthcare around cost per worked hour by unit, by week, with every premium included. Then layer in vacancy cost. It’s less elegant than a single agency number and far more useful.
How staffdna.com Helps With Total Cost of Workforce in Healthcare for Employers & Facilities
StaffDNA was built around the pieces the old model misses. The platform runs a direct-sourcing marketplace, so you see transparent rates with no hidden agency markup layered on top. You know the actual delivered cost before you commit to a req.
Specifically, staffdna.com gives you:
- Real-time rate visibility across travel, per-diem, local contract, and permanent placement in one place
- Float pool and internal resource pool management, so your own staff and external talent sit in the same scheduling view
- Credential and compliance tracking that cuts the administrative hours buried in your fully loaded cost
- Fill rate and time-to-start data by role and market, not just posted bill rates
- Spend reporting that spans contingent and internal labor, instead of treating them as separate budgets
The honest limitation: this works best if you’re willing to change how you source. Facilities that bolt it on beside five existing vendor relationships get less out of it than those who consolidate.
Ready to see your real cost per worked hour? Request a demo at staffdna.com and bring one unit’s last 90 days of scheduling data.
What to Change in the Next Quarter
Recalculate your fully loaded multiplier using current benefits costs. That alone tends to shift the picture.
Then pull overtime by unit for the last 24 months and overlay it against agency hours. If the lines cross, you didn’t reduce spend. You relocated it. Set a target agency range instead of a target of zero, somewhere between 3% and 6% depending on your seasonality, and defend it when finance pushes lower.
Move your labor reporting to a weekly cadence. Monthly is a historical document.
Frequently Asked Questions
What’s the biggest recent change in the total cost of workforce in healthcare?
Benefits and overtime overtook agency labor as the main growth drivers. Agency spend fell roughly 50% from its peak, while employer health costs rose about 9% in 2025 and overtime climbed to absorb reduced contract hours. The total barely moved for many facilities.
Is reducing agency usage to zero still the right goal?
No. Most facilities find that eliminating contract labor entirely pushes them into overtime premiums and higher turnover, which cost more than a controlled 3-6% agency mix. Treat contingent labor as a flexibility tool with a target range.
What fully loaded cost multiplier should employers and facilities use now?
Somewhere in the 1.33-1.40x range for most US acute settings, versus the 1.25x many models still carry. Run your own number using actual 2025-2026 benefits, payroll tax, PTO, and training costs rather than borrowing a benchmark.
How often should we report labor cost?
Weekly, at the unit level. Scheduling decisions happen daily, so a monthly close gives you no chance to intervene. Keep the monthly view for finance and the weekly view for operations.
Does internal float pool labor really cost less than agency?
Sometimes, but the margin is thinner than most business cases claim once you include recruiting, premium differentials, scheduler time, and idle hours during low census. Cost it fully before expanding it.
Conclusion
Key Takeaways:
- Agency spend dropped, but overtime and benefits absorbed the savings, so audit both before claiming a win
- Update your fully loaded multiplier to current benefits reality; 1.25x is now too low for most facilities
- Measure cost per worked hour weekly by unit, with all premiums included, instead of watching agency invoices alone
The old playbook solved a 2021 problem that no longer exists. Rebuild your model around the components that are actually growing, and set a realistic agency target rather than chasing zero. Start with a single unit at staffdna.com and see what your delivered cost really looks like.
