The Float Pool Math. What It Costs to Stop Renting Nurses.

Last Monday I promised you the internal float pool arithmetic. This week: what a pool actually costs to build, the utilization number that decides whether it saves a dollar, and the point where it stops making sense entirely.

By Christina Archer, Founder · Week of August 24, 2026

This year, 70.7% of hospitals told NSI they plan to decrease travel and agency usage. Last year, 73.5% said exactly the same thing. Two consecutive years of near-identical intent, and agency lines that mostly did not move.

The intent is not the problem. Nobody has to be persuaded that renting a nurse at $91.23 an hour is expensive. The problem is that almost every one of those pledges gets built into a budget without anyone running the one calculation that determines whether an internal float pool saves money or simply relocates it.

That calculation is break-even utilization, and it is the entire ballgame. Here it is, on the national averages, in about ninety seconds.

▪The spread is real, and it is the ceiling — not the number

Start with what everyone quotes. A travel RN costs the average hospital $189,758 a year. An employed RN, fully loaded at a 25.8% benefit burden, costs $123,676. The gap is $66,081 per position. Swap twenty travelers for twenty employees and the arithmetic hands you $1.32 million.

That number is accurate. It is also the ceiling, and a ceiling is not a forecast. It assumes the employed nurse is doing productive work every hour you pay her, which is precisely what a float pool cannot guarantee — the whole design premise of a pool is that she is not assigned to a fixed unit.

Convert both sides to hours and the trade gets honest. That $123,676 across a 2,080-hour year is about $59.46 per paid hour, and you carry it whether or not she is deployed. The travel nurse at $91.23 is billed only when she works. Divide one by the other and the answer is 65%.

A float FTE has to be productively deployed roughly two-thirds of her paid hours just to break even against the agency nurse she replaced. Above that line, every point of utilization is margin. Below it, you did not cut agency spend — you moved it onto a different cost center and gave it a better name.

▪What actually pushes utilization below the line

Nobody builds a pool planning for 50% deployment. Four things do it anyway, and they are all design decisions rather than bad luck.

Narrow cross-training. A float nurse cleared for two units has her utilization capped by the combined variability of those two units. Clear her for six and the demand curve smooths out. The pool’s productive hours are a function of how many places she can legally and safely go, which means the orientation budget is not overhead — it is the yield.

Specialty depth. ICU, OR, and L&D floats carry long orientation and cannot efficiently backfill med-surg. The higher the required specialization, the fewer the deployable hours, and the faster the math turns against the pool. Med-surg and tele are where pools earn; high-acuity specialty pools rarely clear break-even on cost alone, and should be justified on continuity and safety instead.

Cancellation culture. A pool that gets cancelled at low census is a pool being charged for idle hours by managers optimizing their own unit budget. That behavior is rational at the unit level and catastrophic at the system level. If the pool is centrally funded and locally cancellable, you have built a machine for destroying your own business case.

Episodic agency use. If your contract labor is seasonal, single-unit, or covering a defined leave, a standing pool costs more than it displaces. The agency contract is the correct tool for lumpy demand. Pools require chronic, distributed, predictable vacancy — which, at a 8.6% national RN vacancy rate and an average of 43 unfilled RN FTEs per hospital, most systems genuinely have. Confirm yours before you assume it. Your own vacancy pattern by unit is in the Workforce Intelligence Center.

▪The failure mode nobody puts in the business case

Here is the finding that should reshape how you price these roles. Nurse.org’s 2026 State of Nursing survey identified float pool nurses as the least satisfied group in the profession — constant reorientation, no team, no unit identity, and the hardest assignments on the worst-staffed nights.

Now look at what the market pays them. Float RN postings average $45.87 an hour against a $47.96 national RN average. The role with the highest adaptive burden in nursing is being posted at a discount to the role with the least. That is not a market signal. That is a design error, and it produces exactly the outcome you would expect.

Price that error. A twenty-FTE pool turning over at the national 17.6% loses between three and four nurses a year at $60,090 per separation — call it $210,000, before you count the orientation you just wrote off and the deployable hours you lost while the seat sat empty for 78 days. A pool built at a discount will churn through your savings faster than the agency line ever did. The retention playbooks that apply here are the same ones in the Nurse Retention Center.

Organizations that get this right do the opposite of what the postings suggest. They pay a real differential, guarantee the schedule instead of cancelling into it, give the pool a manager and a home base, and treat the assignment as elite rather than leftover. One six-site system that built its internal pool that way projected $5 million in contingent-labor savings in a single year. The differential is what buys the utilization, and the utilization is what buys the savings.

▪Three moves before you write the business case

1. Run your own break-even before you run the savings. Take your fully loaded employed-RN hourly cost, divide by your actual blended agency bill rate, and you have the utilization floor your pool must clear. Every projection you present should show the savings at that floor, at ten points above it, and at ten points below. If the downside case is not in the deck, the deck is marketing.
2. Fund cross-training as revenue, not overhead. Count the units each pool nurse is cleared for and treat that number as the yield lever it is. Widening one nurse from two units to five does more for your utilization rate than any scheduling software you are being sold, and it costs a fraction as much.
3. Convert one agency line, not the program. Pick the single unit with the most chronic, most predictable contract coverage. Stand up four to six FTEs against it, guarantee their hours, pay the differential, and measure deployed hours weekly for two quarters. A pool that clears break-even on one unit will scale. A pool that cannot has told you something important for the price of six FTEs instead of sixty.

An agency contract buys you coverage and charges you only for the hours you use. A float pool buys you a permanent asset and charges you for every hour you own it. Neither is the smarter instrument in the abstract — the one that wins is decided entirely by how many of those owned hours you actually put to work. The hospitals cutting agency spend this year are not the ones with better intentions than the 70.7%. They are the ones who ran the utilization number first.

Next Monday: the pay compression audit — why your five-year nurse is the flight risk nobody is tracking, and what the gap between her rate and your new-hire rate is quietly costing you.