Shift bidding for nurses isn't new. Nobody has asked why the open shifts exist.

There is a shift on a hospital unit next Tuesday and nobody is assigned to it. Here is what has been built to solve that.

Float-and-call lists. Shift posting. Internal resource pools. Per diem panels. Float pools. Agency and travel contracts. Incentive tiers. Quick pay. And bidding, in six formats, including software that will keep bidding a nurse’s price down on that nurse’s behalf, round after round, until it reaches a floor they set in advance.

Twenty-five years of real engineering, considerable ingenuity, and a great deal of time, energy, and money, all of it aimed at filling that shift.

Almost none of it aimed at asking why the open shifts exist.

I am writing this now because of a report circulating this week that a large academic health system will move its open shifts to a bidding model this fall. Nurses see a posted shift, bid on it, and the shift goes to the lowest bid.

I have not been able to verify it. The internal document has not surfaced, and every version I can find traces back to the same description repeated on social media from account to account. So this is not about that hospital. It is about the model, which is real, and about the question nobody has asked of it in twenty-five years.

There is a lot of anger about this online right now, and nearly all of it is aimed at administrators. It is aimed one layer too low.

The people making these decisions are working from a number that was calculated correctly, using the method everyone is taught to use. The method is the problem. It is a flawed nursing budget methodology that has been taught and published since 1960, and everyone downstream of it has been deciding with incomplete information and doing the best they could with what they had.

And it is not that nobody noticed. In 1984, in Steven Finkler’s Budgeting Concepts for Nurse Managers, the standard graduate text on the subject, C. M. Graf wrote the chapter on the operating budget and put a footnote on page 73 saying the method described there, while widely used, was not theoretically correct. The chapter printed the right formula beside it, named why the shortfall happens, which is that the positions you add have leave and orientation of their own to cover, and pointed out that the difference adds up to real money once applied across every unit. The same acknowledgment was still there in the second edition eight years later.

And it was corrected again, more directly, much later. In 2016 William J. Ward Jr., a professor of health finance at the Johns Hopkins Bloomberg School of Public Health, set both methods side by side in a table in a budgeting textbook and showed the additive one coming out short by nearly half an FTE on a seventeen-position example. The warning is blunt: the approach “is clearly incorrect and will always result in a budget that lacks sufficient staff.” That book is not about nursing. It is about hospital budgeting generally.

The correction has been sitting in the textbooks the whole time. It just never took.

If you want something to be angry at, that is the better target than the person who posted the shift or who authorized the purchase of a shift bidding system.

The alarm assumes nurses are bidding their wages down with no floor. That is not quite what these systems do, and the difference matters if you are the one bidding.

Shift bidding applies to extra shifts. It is work nurses pick up beyond the schedule they already committed to, so the base rate for regular hours is not what is in play. Shifts are posted with an hourly maximum and a minimum, or sometimes as a percentage range above base pay. There is a floor. Credentials gate who can bid at all, so only qualified staff see a given shift. Some systems award the shifts automatically to the lowest qualified bid, and others leave the decision with the manager who posted it.

It is also not only registered nurses. When Sharp HealthCare in San Diego brought its system in, nursing assistants got it first, as the trial group, and RNs followed months later. The 2005 buyer’s guide describes the unit budget these shifts come out of as covering registered nurses, licensed practical and vocational nurses, nursing assistants, and ancillary staff. Treating this as an RN story leaves out part of who is actually bidding, and part of who is short.

None of which makes the concern silly. A descending auction on clinical labor deserves scrutiny. So does the reason there are that many open shifts.

St Peters Hospital in New York was running its own shift-bidding system in 2001, and by 2004 had put 127,000 shifts through it. Sharp, which runs seven facilities and employs more than 3,500 registered nurses, published its own implementation account in Nurse Leader in August 2004 and named two other systems already doing the same thing. One of them reported cutting its use of outside temporary staff by more than ninety percent.

The vendors are equally old. BidShift was a new commercial product when Sharp bought it, built jointly with a health system in Indianapolis. In September 2005 the California HealthCare Foundation published a buyer’s guide to these systems for hospital executives, cataloging six bidding formats. One of them it calls a reverse auction. Another is the proxy bidding described above, which by then was a feature you could shop for.

That was twenty-one years ago, in a document written to help hospitals choose a vendor. The software Sharp bought has been renamed and changed owners since, and shift bidding is still on the market.

When Sharp brought its system in, staff and the union accepted it on a condition: only shifts that the unit’s own staff did not want to fill would go to auction. The vendor then built a feature that could restrict an auction to a single unit or a defined group of staff for a period before opening it to everyone else.

So if your employer is considering this, there are questions to ask at a staff meeting, and they have answers. Which shifts are eligible. Whether there is a floor, and how low it goes. Who gets first refusal. Whether the award is automatic or a manager decides.

Ask them. The answers decide whether the thing is livable day to day, and that is not a small matter.

They do not settle everything, though. Those questions cover which shifts and which people. There is a separate question about how the price gets set, and that one is harder.

Premium pay and a descending auction get discussed as if they were the same tool. They are not, and the difference is the direction.

A premium moves the price up. The shift stays open, the employer raises what it will pay, and it keeps raising until somebody takes it.

That is not a clean process either, and anyone who has worked a floor knows it. Nurses who understand the pattern wait. The rate climbs as the shift gets closer and the staffing office gets more anxious, so holding out is rational, and plenty of people hold out.

A descending auction moves the price down. Staff underbid one another, and where the award is automatic, the shift goes to the lowest bid.

Both are bargaining over the same gap. The difference is which side can afford to wait, and so who ends up absorbing the cost of a shift that should never have been open. When the price climbs, the institution pays for its own shortfall. When the price falls, the nurses do.

Two nurses, same license, same unit, same shift, same work, at different rates. What separated them was not skill, seniority, or how hard the shift was. It was which one was willing to undercut their colleagues to be certain of the shift.

The nurse who most wants the extra work is the one who earns the least for doing it.

It also turns colleagues into competitors over the price of their own labor, on a unit where they will be working alongside each other the next morning. Teams that have to hand off patients to one another are not helped by having bid against one another for the work.

A floor restricts how far that can go. It does not change the direction it travels.

And there is something strange about building this particular mechanism into a market where the employer’s own position is that it cannot find enough people. Where a buyer is short of what it wants, prices are supposed to rise. A descending auction does the opposite. It is a device for making the final price fall as more people compete for the work.

Which brings me back to the shift on Tuesday that nobody is assigned to.

An open shift is the gap between what the schedule needs and what the unit’s budgeted positions cover. That position count is a number. Somebody calculates it once a year. It is the input to every mechanism at the top of this piece, and it is the one input none of them examines.

There is a documented way for that number to come out too low.

The most common budget approach in nursing finance education starts with the hours nurses spend delivering patient care, then adds a percentage on top to cover the rest: orientation, education, time off, and leave, the time that supports the care without being at the bedside. Call those two things Patient Care and Patient Care Support.

The trouble is that the support time is a share of the whole staffing requirement, not an addition to the direct portion. Take a unit that needs one hundred full-time positions, eighty of them Patient Care and twenty Patient Care Support. Start from the eighty and add twenty percent, and you get ninety-six. Divide the eighty by 0.8, which is what the relationship actually is, and you get one hundred.

Four positions. They were never funded, so they were never posted, never vacant, and never counted anywhere. They show up instead as a shift with nobody assigned to it, one day at a time, all year.

And look at what those four were for. The twenty was the coverage for sick calls, vacation, orientation, education, and leave. Fund ninety-six instead of one hundred and you have the eighty delivering care and sixteen of the twenty covering absences at the bedside. The budget named the coverage, priced it, and then funded a staff that cannot supply it.

So when somebody calls out on Tuesday, the unit is short of exactly the person the budget said should be there. The shift goes up for bid, and the people bidding on it are covering an absence the plan already accounted for and did not pay for.

For a sense of scale, one California hospital reported filling more than 11,000 direct-care hours from its own staff in the first three months after switching its system on. That is the volume of work that fell outside the resources the hospital had available. One hospital, one quarter.

It does not tell you the cause. A unit can be fully funded and still short on a given Tuesday, through a combination of vacancies, orientation, time off, and leave. But those are not a rival explanation. They are the coverage the budget is meant to fund, and they are the part the common approach understates. And the positions the arithmetic leaves out never show up in a vacancy report at all, because a position that was never funded cannot be unfilled. So the real gap is the reported vacancy rate plus another two to six percent of what the unit should have had. Where it falls in that range depends on how much support time the budget assumed.

A finance officer will point out that this corrects one step, and that the direct care hours going into it are their own argument. That is fair, and it does not rescue the calculation. Take a unit that has already settled on a one-to-four nurse to patient ratio and a fifteen percent allowance for support time. Both are the institution’s own stated assumptions, decided before any of this arithmetic starts. One of them gets applied correctly and the other does not. Whether the ratio itself is right is a separate argument, and if it is wrong, the position count is wrong twice rather than once.

I cannot tell you how any hospital named in this piece calculated its positions. I do not know. What I can tell you is that the approach that understates is the one that runs through nursing finance education, and has for decades. I have documented instances of it across textbooks, continuing education, and the peer-reviewed practice literature. The earliest is a 1960 monograph published by the Catholic Hospital Association. The same arithmetic appears in a 1987 business skills text for nurse managers from J.B. Lippincott, and again in a nursing administration textbook published in 2020. Sixty years, three publishers, one method. So the odds that none of these budgets was built that way are not good. That is an inference from how widely the method has been taught, not a finding about any particular hospital, and the reason I cannot do better is that nobody checks.

If the budgeted positions are right, some of this machinery is doing ordinary work, and census moves and sick calls are real. If they are wrong, the auction is pricing a gap the institution manufactured, and running it efficiently makes the error cheaper to carry indefinitely. Nobody knows which, because the number does not get audited.

There is also an industry on the other side of that gap. Open-shift platforms, per diem panels, float-pool software, agency and travel intermediaries: each is a business whose revenue moves with the number of shifts the budget did not cover. In 2005 a stand-alone product that let nurses bid on open shifts ran three to nine thousand dollars a month for a three-hundred-bed hospital, and a full scheduling and staffing system ran sixty to a hundred and fifty thousand. None of these companies set out to profit from an arithmetic error, and I do not think any of them knows it is there. But none of them has a reason to want the gap closed either, and the correction is the one thing none of them can sell.

Three consequences that show up in the numbers, not only in morale.

They are also the three arguments that travel in a budget meeting. If you ever find yourself in one, these are the ones to have.

A descending auction selects on the wrong variable. The shift goes to whoever bid lowest, so what the mechanism selects for is willingness to accept less. Not fit for the unit, not familiarity with those patients, not the right skill for that acuity. In clinical work, that is not a quality-neutral way to decide who is at the bedside.

It degrades the measurement. The success metric in every published account of these programs is the vacancy rate, and it moves impressively: eleven percent down to five at one hospital, twenty down to seven at another. Those are real achievements, driven by retention and by pulling agency nurses onto staff. But a vacancy rate is unfilled budgeted positions divided by budgeted positions. Both numbers come out of the budget. Every one of these programs improved the ratio, and not one of them put the denominator in question.

It makes the defect affordable. A cheaper way to fill gaps lowers the pressure to fix whatever is producing them, and an auction is cheaper than a posted premium precisely because it takes back the advantage a nurse gets by waiting. If the position count is short, an efficient auction is a subsidy for the shortfall.

There is a fourth, and the industry found it immediately. Sharp had to rewrite its payroll system because nurses already in overtime were automatically paid their overtime rate instead of the bid rate they had won. The 2004 Nurse Leader account calls accommodating federal overtime law a daunting task. Variable rates for identical work are a compliance question before they are anything else, and that was clear in 2004.

The people who built these systems solved a real problem. Filling a shift at eleven at night is a genuine job, and doing it by phoning down a list was worse for everyone. The hospitals that adopted them were losing nurses to agencies that offered more money and more control, and they responded by offering their own staff more money and more control. They were also watching their own incentive rates climb every time a shift went unfilled. Nurses got access, flexibility, and higher earnings. Managers got their evenings back.

That includes the nurse managers, who take more of the blame for this than they have earned. A manager does not set the position count. It arrives. Somebody hands them a number and a schedule to cover with it, and when the number will not cover the schedule, the manager reaches for whatever tool the organization put on the desk. Usually that means overtime, agency staff, or the resource pool. At some places it now means an auction. Being handed a short staffing plan and a cheap way to fill the gap is not the same as choosing either one.

Every one of those people answered the question in front of them. The question was how to fill this shift. It was a reasonable question and they answered it well.

It was just never the first question.

The direction did get decided, though, whatever else is true. Sharp found that the unit-level incentive plans it already had were competing with the bidding system, and reported that it was working to replace those programs with bidding instead. A posted premium that nurses can wait out costs the institution money. An auction does not. Moving from one to the other is a reasonable thing for a hospital to do, and it is a decision about who holds the better position, taken twenty years ago and never argued in public.

But look at what everyone is fighting over. Nurses hold out for a better rate. Hospitals build machinery to stop them holding out. Both sides are bargaining hard, and skillfully, over a shift that exists because the position that should have covered it was never funded. The rate fight is real. It is also a proxy for a number that one side of the fight sets and the other never sees.

Check the number.

Take one unit. Find how its budgeted positions were calculated. If the support time was added as a percentage on top of the direct care hours rather than treated as a share of the whole, the unit has been running short since the day the budget was approved, and no amount of clever shift-filling will fix that, because the machinery downstream was built to accommodate the gap rather than to close it.

If you are not the person who can pull a budget, you can still ask the question. How were this unit’s positions calculated, and was the coverage time added on top or calculated on the whole? Your manager either knows or can find out, and nothing about that question is off limits.

The obvious objection is the cost, and anyone in a finance office will raise it immediately. If the number is wrong, the correct number is bigger, and a bigger number means more funded positions. That is real money.

The money is already going out. It leaves as overtime, as agency premiums, as bid rates, and as the recruitment and orientation costs behind a unit that runs short all year. Sharp expected to save more than a million dollars in 2004 mostly by pulling that spending back in house.

And the labor line is the smaller half. Across decades of research, nurse staffing levels are associated with increased mortality, failure to rescue, readmissions, and length of stay. None of those are soft costs. Length of stay is a direct one, and readmissions and hospital-acquired conditions carry payment penalties a finance department already tracks by name. A 2026 study of eighty-two medical and surgical units across nine hospitals found understaffed shifts associated with higher mortality, more readmissions, and longer stays. An economic evaluation covering four English hospitals and more than six hundred thousand patients found that eliminating understaffing produced net savings once the shorter stays were counted, and work in Illinois found the savings from better staffing exceeded what the additional nurses cost.

Which points at the uncomfortable part. The benefits of fixing this land on patients and nurses. The costs land on the hospital. That asymmetry is a better explanation than indifference for why twenty-five years of evidence has not moved the number.

The question is not whether to pay for the coverage. It is whether to fund it once, deliberately, or buy it back at a premium every week for a decade.

If the number holds up, good. The tooling above it is doing honest work and should be judged on its own terms.

But somebody has to look. Twenty-five years of increasingly sophisticated systems have been built on top of that figure, and in all that time it has been treated as the one thing not worth verifying. It is cheap to check. Nobody is checking.

Davis A, Athis A, Douglas K. “Pioneering New Solutions in Nurse Staffing: Implementing a Bidding System for Filling Open Shifts.” Nurse Leader, August 2004. Sharp HealthCare’s own account, and the source for the 2001 St Peters date, the Spartanburg figures, and the overtime problem. One of its authors worked for the vendor, which matters when reading it.

Sabet L. Adopting Online Nurse Scheduling and Staffing Systems. First Consulting Group for the California HealthCare Foundation, September 2005. The buyer’s guide that catalogs the six bidding formats, including the reverse auction and proxy bidding.

Ellerbe S. “Staffing through Web-based open-shift bidding.” American Nurse Journal, April 2007. Tri-City Medical Center, and the source for the 11,000 hours.

On the vendor history: socaltech.com on the Concerro renaming, and the API Healthcare announcement of the Concerro acquisition, 8 February 2012.

Graf CM. “The Operating Budget.” In Finkler SA, ed., Budgeting Concepts for Nurse Managers, 1st ed. W.B. Saunders, 1984, pages 62 to 103. The 1984 correction; the footnote is on page 73, and it survives in the 1992 second edition.

Ward WJ Jr. Health Care Budgeting and Financial Management, 2nd ed. Praeger, 2016, chapter 6, pages 128 to 130. The 2016 correction. Table 6.14 sets the two methods against each other: 20.00 FTEs calculated correctly against 19.55 calculated additively, a shortage of 0.45.

On the outcomes and the economics: Morioka N, Moriwaki M, Miyawaki A, Saville C, Fushimi K, Griffiths P. “Hospital Nurse Understaffing and Patient Mortality, Readmission, and Length of Stay.” JAMA Network Open, February 2026. Saville C et al., the economic evaluation of eliminating understaffing across four English hospitals, discussed in Lasater KB, “Eliminating hospital nurse understaffing is a cost-effective patient safety intervention,” BMJ Quality & Safety, 2025. Lasater KB, Aiken LH, Sloane D, French R, Martin B, Alexander M, McHugh MD. “Patient outcomes and cost savings associated with hospital safe nurse staffing legislation.” BMJ Open, December 2021.

On the budget method itself, I keep a running list of published instances. The earliest is from 1960. They run through textbooks, continuing education, and peer-reviewed journal articles, and they continue into the present.

© 2026 Robert L Wingo - All rights reserved.

Robert Wingo, BSN, RN, NI-BC is the author of the Nurse Staffing Information Structures (NSIS) framework, a mathematical and conceptual approach to understanding and describing the nurse staffing life cycle from reimbursement through outcomes. He is a board-certified nursing informatics specialist, has worked in nurse staffing and scheduling operations, healthcare finance, and data analytics since 1999, and is a Nursing Economics Fellow (FINE Fellowship, Commission for Nurse Reimbursement).