Healthcare Workforce Planning: The $10M Wage-Mix Gap in Your RN Budget

The Wednesday-night scene

It's 6:47 PM on a Wednesday. Sarah, VP of Finance at a 4-site community health system, is looking at the FY26 nursing budget. Her CNO's email is open in a second window. Three sites are running 45% agency staffing this quarter. Base budget assumed 100% permanent. The audit committee meets Friday.

The position roster came over from HR: 312 RN FTEs at an $89,000 blended base. Sarah's predecessor built the model. It has 14 tabs, cross-linked formulas, and a broken reference from a spring column rename that nobody caught until this afternoon. Total-comp rolls up to $56.2 million on the summary page. That's the number the CFO showed the board last quarter. It's also the number Sarah now believes is understated by somewhere between $8 million and $12 million, because the agency mix isn't modeled at all. It's assumed away.

Every Wednesday at 6:47 PM there is a version of Sarah at every 2-8-site health system in the country. She's looking at a spreadsheet that was accurate when it was built, that has drifted since a nurse-manager reorg, and that carries an assumption (100% permanent) that stopped being true two quarters ago. The auditors won't catch it. The CFO won't catch it. The actuals will, and by Q2 the variance will be a board conversation.

That moment is the one this piece is about. This is written for the Controller and the VP Finance who are quietly aware that the model is drifting, that the workarounds are stacking, and that a real answer to "what will nursing cost us next year?" is beyond what Excel can carry.

The rest of this piece is the math, the mechanic, and what to do about it Monday morning.

‍

Why your model is wrong at position, wage-mix, and fringe

Three structural realities are true for every 2-8-site health system finance team. Each one breaks Excel a little differently, and together they compound.

Position level, not employee level. While corporate functions like finance and HR may budget neatly by individual employees, clinical teams face a different beast. With high turnover rates, the roster changes often. The name on the spreadsheet row is just a placeholder; what matters is the position itself. Many healthcare finance teams resort to creating parallel Excel models to deal with this mismatch, and we all know how messy that can get.

The geometric explosion. Take 4 sites, 3 RN role types (staff RN, charge RN, specialty), 3 wage tiers (permanent, travel, agency), 4 benefit tiers, and payroll-tax variation by state. That's 4 × 3 × 3 × 4 × 2 = 288 cell combinations for RN staffing at a single hospital system. Add allied health (respiratory, imaging, lab, therapy). Add non-clinical. You're modeling 2,000+ line items to get one number: total labor. Excel fails quietly here. A missed benefit tier on one site rolls up as a 3% understatement of total-comp, which the CFO then carries into the board deck as the operating margin projection.

The four-layer fringe ladder. Every clinical FTE carries four separate compensation costs beyond base wage. PTO accrues by hours worked, not by headcount. FICA and Medicare are fixed percentages, but the wage-mix change moves the base they apply to. Disability and workers' comp vary by role class (an OR nurse and a floor RN price differently). Bonus tiers vary by group. One healthcare finance operator we spoke with described her bonus structure as "7% or 10% or 20% by group, and being able to allocate to different departments." Each layer composes with the FTE unit differently. Miss any one and the total-comp forecast drifts 8-12% from actuals over a fiscal year. Enough to invalidate the operating margin conversation with the board.

Excel can hold each of these individually. It can't hold all three at once at position-level detail across 2-8 sites without a Controller full-time on model maintenance. Which is what actually happens. A person whose job description does not say "Excel administrator" is spending 60% of their time on model administration.

Sit with what "60% of a Controller's time" costs the business. A Controller at a $50M revenue health system runs $110,000 to $145,000 all-in. 60% of that time on model maintenance is $66,000 to $87,000 of annualized Controller labor pointed at Excel plumbing instead of variance analysis, capital-project modeling, or the questions the CFO actually wants answered. And that's the visible cost. The invisible cost is that the Controller can't take a two-week vacation without the budget model going stale, because the model is a set of formulas and rules that only she knows. It's a person-shaped dependency, and it's on your risk register whether you've written it down or not.

‍

Working through the wage-mix and fringe math at 3,000-FTE scale

Here's a worked example at healthcare scale. 4 sites, 3,000 total FTE (roughly the Kindera Living scale, see the customer stories below). Mix of clinical (nursing plus allied health) and non-clinical (finance, HR, IT, facilities). Enough complexity to be realistic, small enough to walk through end to end.

Step 1. Position-level roster. For nursing, build the model at position, not at employee. Site 1 has 42 RN positions across 3 role types. Site 2 has 58. Site 3 has 39. Site 4 has 51. Total: 190 RN positions before you touch turnover or vacancy. Each position carries a target FTE (1.0 for full-time, 0.5 or 0.75 for part-time), a target wage band, and a target benefit tier. The name in the row is metadata. The forecast holds when the person turns over.

For non-clinical roles, keep the employee-level rows. Finance staff, IT, HR, facilities: 8% turnover, low mix volatility. Employee-level is fine.

Structural point: your FP&A tool has to hold both models in one workbook. If it forces you to pick one, you're back in the Excel-hack pattern.

‍Step 2. Wage-mix layering. For nursing, split each position's forecasted hours across three tiers:

• Permanent (staff RN on payroll, at your posted wage bands, with full benefits)
• Travel (contracted through a travel-nurse agency, at negotiated rates, no benefits from you)
• Agency (spot-labor, the highest-cost fill for shortfalls, no benefits from you)

The mix is the number that moves. Base budget usually assumes 100% permanent. Every hospital finance team knows that's wrong. Nobody models the alternative because doing so in Excel means duplicating the position roster across three tabs and reconciling.

Here's the swing. At the 3,000-FTE scale, with 190 RN positions and an average blended base of $95,000 including benefits at 100% permanent, budgeted RN cost is roughly $18 million. Take the same 190 positions at 40% permanent, 40% travel (at $135/hour, 2,080 hours, no benefits), 20% agency (at $195/hour, 2,080 hours, no benefits). Actual RN cost lands closer to $28 million. A $10 million gap on one role type at one system. The gap is real, it shows up in the ledger every month, and it is not in the budget the board approved.

Step 3. Fringe ladder composition. For the permanent-tier hours, layer the four fringe components:

• PTO accrual: 12% of base wage for RNs (benchmark for a 4-week vacation policy accrued by hours worked)
• FICA and Medicare: 7.65% of gross wage (federal, fixed)
• Disability and workers' comp: 4% to 8% of wage depending on role class (an OR nurse prices higher than a floor RN)
• Bonus tiers: 7% or 10% or 20% by group, allocated to the departments where the work drives the accrual

Total fringe on permanent RN hours: 30.65% to 47.65% of base wage. That's a range, not one number. It depends on which role and which bonus group. Get the range wrong on the average and the total-comp forecast drifts $2-4 million for RN alone.

Step 4. Payroll-tax allocation across cost centers, within a single entity. The final composition step: roll the loaded total-comp per position up to a cost center, then to a division, then to the operating entity. Payroll taxes (state unemployment, federal unemployment, state disability where applicable) allocate by wage base, not by hours. So you need the wage base per site (state matters), the tax rate per site, and the allocation rule per cost center. A 4-site system spanning three states carries 12 tax rate combinations before you touch anything else.

Content-uniqueness note: this piece stops at single-entity roll-up. Cross-entity payroll allocation, where an operating company, a 501(c)(3) foundation, and separate housing or clinical entities share a chart of accounts, is a different problem. See the companion blog on multi-entity healthcare consolidation.

At the end of steps 1 through 4, you have a defensible total-comp forecast: position by position, tier by tier, fringe layer by fringe layer, cost center by cost center. Every cell of the summary traces back to a specific input. Change one input (bump agency mix from 20% to 25%, or add a new bonus group), and the forecast updates in one place, not fourteen.

The defensibility matters more than the forecast. When the auditor asks "why is nursing total-comp $28 million and not $18 million?", you can answer in three sentences: mix moved from 100% permanent to 40/40/20 permanent/travel/agency during Q2, the travel and agency tiers price at $135 and $195 an hour with no benefits burden on us, and the composition of the fringe ladder shifts because those tiers carry no PTO, FICA, or bonus accrual. The auditor will follow the trace back to the input cell, verify it against your Kronos or Kronos-equivalent time-keeping data, and move on. That is what the Controller's 60% of time has been buying you all along, one variance at a time, one call at a time. The model just does it now, and the Controller can spend her Q3 on the capital plan.

‍

What Centage customers actually do at this scale

Three named healthcare finance operators talk about this workflow on the record. Their stories cover the pain from three angles.

‍Kyle Raeder, Director of Reimbursement & Financial Planning, Community Care, Inc. Community Care is a non-profit health and human services provider. Multi-provider payroll allocation across community health, hospice, and behavioral care. Kyle's team faced position × cost-center × cross-provider complexity at a scale where the Excel workaround was consuming a full FTE of Controller time.

His verbatim:

‍"Centage gave us the flexibility we knew we needed, but thought we couldn't achieve. It allowed us to totally change our approach to the budgeting process."

The read: finance-led migration, no IT team required, budget cycle contracted, model became defensible to auditors and the board. "Totally change our approach" is the operative phrase. It's a different workflow, owned by finance, that produces a number the CFO can stand behind.

Monica Fernos, VP of Budget & Financial Analysis, Medical Card Systems. MCS is a healthcare insurance operator with 2,400 employees and roughly $380M in operating expenses. Enterprise-scale complexity, including the four-part fringe ladder discussed above.

Her verbatim:

‍"We just upload. There are so many fringe benefits, vacation, Medicare, disability, bonuses. If I had to do that manually for 2,400 employees, it would take me a year."

The scale is what makes the story land. At 2,400 employees, manual fringe math is a full-time role that Monica's team doesn't have. Her follow-up on granular reporting to the CFO and CEO speaks to the second-order value: once the math is automated, the finance team can spend their time on interpretation. Which is the job the CFO actually hired them for.

Angela Groza, Executive Vice President of Finance, Kindera Living. Kindera is a long-term care and retirement organization with 3,000 employees across the organization. They tried the bundled-planning-module option first (NetSuite Planning) and came back to purpose-built FP&A.

Her verbatim on why the bundled option failed:

‍"We spent a considerable amount on the implementation. But ultimately, it was an expensive lesson learned."

The context matters. NetSuite Planning is a real product from a real vendor, and it works for some organizations. For Kindera's position-level workforce depth (they budget 3,000 employees at position, not employee, to keep the model clean and manageable), the bundled tool couldn't reach the workflow they needed. Angela's operational summary: "We have 3,000 employees across the organization, but we budget at a position level, not at the employee level. It keeps our budget clean and manageable while still giving us the detail we need."

That's the workflow this piece has been walking through, in one sentence, from someone who runs it every month.

Three named operators, three angles: finance-led migration (Kyle), fringe math at enterprise scale (Monica), position-level depth after a bundled-tool failure (Angela). Every one is a Centage customer today, and every one talks about workforce planning as the workflow that changed.

The pattern across the three stories is easier to see when you line them up against the four steps of the mechanic. Kyle's team hit the wall at Step 4, payroll-tax allocation across cost centers, because Community Care's chart of accounts crosses community health, hospice, and behavioral care with distinct payroll rules per provider. Monica's team hit the wall at Step 3, fringe ladder composition, because 2,400 employees with four fringe layers is $10 million to $15 million of composition math per fiscal year that no human should be doing by hand. Angela's team hit the wall at Step 1, position-level vs employee-level roster, because 3,000 employees at employee level was a 3,000-row spreadsheet that turned over 12% a quarter, and NetSuite Planning couldn't reach position-level depth without custom code they'd have to maintain. Different step, same failure mode: Excel plus a bundled tool that doesn't fit runs out of runway somewhere between step 1 and step 4. What Centage does for all three is hold the composition together in one workbook, so a change at step 1 (adding a position at a new site) flows through step 2 (wage-mix tiering), step 3 (fringe layering), and step 4 (payroll-tax rollup) automatically. The Controller changes one input; the model updates in one place.

‍

What to do Monday morning

You should audit your current model before deciding what to do about it. Four checks, in order.

1. Is your clinical roster at position or at employee? Open the RN tab in your budget workbook. Count the rows. If the row count roughly matches your active-payroll RN headcount, you're at employee level. That works when turnover is under 10% and the roster is stable. It doesn't hold at 40-60% RN turnover. Move to position. Every position gets a target FTE, a target wage band, a target benefit tier, and a name field that's metadata, not a driver.

2. Are all four fringe layers in the model? Search your workbook for PTO, FICA, disability, and bonus. If any of the four is missing, or if any is a single hardcoded percentage across all roles, the total-comp forecast has systematic error. The direction of the error is knowable, and it compounds across the fiscal year.

3. Do you have a wage-mix variable, or are you assuming 100% permanent? If your budget starts with FTE count times a blended wage, you're assuming 100% permanent. Every hospital finance team knows that assumption is wrong. Model the three tiers explicitly. Even a rough split (say, 60/25/15 for permanent/travel/agency) that reflects last quarter's actuals will land closer than the 100%-permanent baseline.

4. Run the swing calc for yourself. Download the Multi-Location Nurse Wage-Mix Worksheet. Plug in your site count, your wage bands, your current mix, your fringe assumptions. The sheet computes the delta between a 100%-permanent assumption and your actual mix. If the delta is under 5%, your model is close and the current Excel workflow probably holds through this budget cycle. If it's 5-15%, your budget has material variance risk and you should invest in a workforce-planning-capable tool before next budget season, because the drift will compound quarter over quarter. If it's over 15%, your budget is structurally wrong for FY26 and the Controller is likely already carrying the knowledge in her head. That's a board conversation, and it should happen in this quarter, not next.

‍

Download the Multi-Location Nurse Wage-Mix Worksheet (Google Sheets, also .xlsx). Plug in your numbers. See what your current model is missing. Email gate. Immediate download.

‍

If this shape of problem is your Wednesday-night reality, we should talk.

Centage's Personnel Module handles position-level wage-mix, the four-layer fringe ladder, and payroll-tax allocation across cost centers natively. 4-6 week implementation. Finance-owned. Onshore Customer Success Manager who understands accrual accounting. A 30-minute demo starts with your actual model.