Shift Swap Policy: Approval, Overtime, and Fairness

The first swap that cost me money was arranged over text message on a Wednesday evening, between two people I liked, for a reason I would have approved instantly if anyone had asked. One of them had a family thing on Saturday. The other wanted the hours. They sorted it out between themselves and told nobody, because from where they stood there was nothing to tell.

I found out the following Friday, from the payroll run. The trade had moved eight hours out of one person's next week and into their current one, which put them at forty-eight, which meant eight hours of overtime on a business that had bought exactly the same amount of labour it always did. Nobody had done anything wrong. The schedule had simply been rearranged in a direction the overtime rules do not forgive.

A shift swap policy is not bureaucracy for its own sake. It is the short list of checks that keep a trade genuinely free: no unplanned premium, no unqualified person on a shift that needs a certificate, no minor over their hours, and a record proving the change came from the employee. I build the people and records side of this at FirstHR. This is general information, not legal advice.

What a Shift Swap Actually Is

A shift swap is a two-way trade: two employees exchange assigned shifts with each other, both keep working, and the total number of hours the business buys does not change. That last clause is what makes it different from coverage, where one person needs to be replaced and somebody else has to absorb the gap.

The distinction is not academic. Coverage is one-directional and usually reactive: someone cannot work, and you need a body. What this guide covers is the trade itself and the wage and hour consequences that ride along with it.

Three things make swaps worth allowing. They solve scheduling conflicts without a manager spending time on them, they give hourly staff a form of flexibility that costs nothing, and they cut no-shows, because a person with a way out on Wednesday is far less likely to simply not appear on Saturday. In a business without a dedicated HR person, that last effect alone justifies the policy.

The catch is that a swap moves hours between weeks as easily as it moves them between people. Two employees trading like for like inside the same week is usually neutral. A trade that reaches across your workweek boundary, or that stacks an extra shift onto someone already near forty, is where the money appears. That is what the rest of this guide is about.

The Approval Workflow, Start to Finish

A working swap process has six steps and takes a manager about ninety seconds per request. The order matters, because the two checks that cost real money happen before the approval rather than after the shift is worked.

Set one hard deadline and defend it. Requests submitted at least a fixed number of hours before the shift, twenty-four or forty-eight depending on how far ahead you plan, get the full process. Anything later stops being a swap and turns into a coverage problem, which follows a different route.

Why Approval Cannot Become a Rubber Stamp

Manager approval has to stay mandatory because it is the only point in the process where anybody looks at the consequences, and it stops being worth anything the moment it turns into a reflex. An approver who says yes to everything within four seconds is not approving, they are logging.

The reason this matters more than it sounds is that the two employees involved genuinely cannot see the problem. Neither of them knows the other's weekly total. Neither is tracking your workweek start. Neither knows whether the Saturday shift requires the alcohol service permit that only one of them holds. They are optimising their own calendars, which is exactly what you want them to do, and the checks are yours to run.

Two habits keep approval real without adding load. Give the approver a decision deadline, so the check has to be quick, and give them a default answer for the ordinary case: like-for-like trades inside the same week between two qualified adults are approved, everything else gets thirty seconds of attention. That way the exceptions are the only thing consuming judgement.

Track your approval rate. If it sits at one hundred percent for a quarter, either your team is unusually disciplined or the check is theatre. If it sits below seventy percent, people will stop asking and start trading privately, which is the outcome the policy exists to prevent.

The Overtime a Swap Creates, With the Arithmetic

A swap can create overtime even when nobody works an extra hour, because overtime is measured one workweek at a time and a trade can move hours across the boundary between two of them. The Fair Labor Standards Act takes a single workweek as its standard and does not permit averaging hours over two or more weeks (29 CFR 778.104).

Here is the case that catches people. Your workweek runs Monday through Sunday. Maya and Chris are both non-exempt and both paid $20 an hour. Maya is scheduled five eight-hour shifts this week, so forty hours. Chris is scheduled four, so thirty-two. Chris asks Maya to take his Friday shift this week, and in exchange he will take one of hers next week. Both of them think of it as even, because across the fortnight it is.

Follow the money. Maya works forty-eight hours in week one, so the first forty are straight time at $20 for $800, and the remaining eight are overtime at $30 for $240, giving her $1,040. Chris drops to twenty-four hours and $480. Week one costs $1,520 against $1,440 before. Week two reverses their hours and lands back at $1,440.

Across the fortnight the business paid $2,960 instead of $2,880 for exactly 144 hours of work. The extra $80 is the half-time premium on eight hours: $10 on top of each. Nobody gained an hour, nobody lost one, and the labour delivered was identical. The calendar did the whole thing.

Two refinements matter in real payrolls. The overtime rate follows the receiving employee's own regular rate, not the rate of whoever gave up the shift, so a trade across different pay rates moves the cost both ways. And the regular rate is not always the base wage: shift differentials and non-discretionary bonuses earned that week fold into it.

The check that prevents all of this takes five seconds and belongs before approval: what is this person already scheduled for in this workweek, and does the shift they are picking up push the total past forty? Exempt employees are outside the calculation entirely, which is worth knowing before you apply the rule to the wrong person (exempt versus non-exempt).

Nobody Can Waive the Premium, Including the Employee

An employee cannot agree to take a swap at straight time, and you cannot accept the offer even when it is made sincerely and in writing. Overtime under federal wage and hour law is not a negotiable term of employment, and an agreement between employer and employee that conflicts with the statutory requirement is simply ineffective.

The regulation is blunt about it. Parties cannot lawfully agree that a rate will be lower because the work is performed during statutory overtime hours, and since a lower rate cannot be set for overtime hours, they cannot agree that the time will not be paid for at all. An announcement that overtime will not be compensated unless authorised in advance does not impair the employee's right to be paid for work the employer suffered or permitted (29 CFR 778.316).

There is a related trap in the phrase authorised in advance. A rule that says unapproved overtime will not be paid is unenforceable as written. You may discipline an employee for working hours you did not approve, and you may make unapproved swaps a policy violation, but you must still pay for the time. Separating the two is the whole trick: pay the hours, then address the conduct.

The practical version of all this is that the only lawful control over swap-driven overtime is the approval gate. Once the shift has been worked, the money exists. Everything downstream is either paying it or building a wage claim.

Predictive Scheduling: Employee Initiated Only If It Really Is

In jurisdictions with predictive scheduling or fair workweek laws, a genuine employee-initiated swap is generally exempt from the predictability premium that a late employer change would trigger. The exemption is real, it is written into the statutes, and it is conditional on the change actually originating with the employee.

Oregon's law requires the written work schedule at least fourteen calendar days ahead and attaches compensation to employer changes made inside that window, while changes resulting from an employee's written request are not subject to the advance notice requirements (Oregon Bureau of Labor and Industries). The statute itself, ORS 653.455, lists mutually agreed employee-initiated shift swaps or coverage among the exceptions to the compensation requirement, and permits the employer to pre-approve them.

Seattle takes the same shape. Its secure scheduling ordinance provides that the additional compensation requirement does not apply to mutually agreed work shift swaps or coverage among employees, and states that the employer may require that it pre-approve those arrangements and may assist employees in finding them. Both jurisdictions draw the same line: helping people find each other is fine, arranging the trade yourself is not.

That line is where employers lose the exemption. Three moves convert a swap into an employer change. A manager who proposes the trade. A manager who tells a specific person to take it. A manager who applies pressure, including the soft kind, to someone who would rather not. In each case the change originated with you, inside the notice window, and the premium can attach exactly as if you had rewritten the schedule.

Thresholds, covered industries, and the size of the premium vary by jurisdiction, and the list of cities keeps growing. Check what applies where you operate using the guides to predictive scheduling laws and schedule changes. Then write the documentation habit into the policy, because in a dispute the exemption is worth precisely as much as your evidence that the employee asked first.

Who May Cover Whom

Not everyone can take everyone else's shift, and the policy has to say so before somebody proves it the expensive way. A swap is only valid if the receiving employee can lawfully and competently perform the work that shift requires, which is a narrower question than whether they are on the roster.

The obvious layer is credentials. Alcohol service permits, food handler cards, forklift or equipment sign-offs, professional licences, security clearances, and vehicle authorisations are all shift-specific, and none of them travel with goodwill. If a Saturday night bar shift requires a permit and the person picking it up does not hold one, the shift is uncovered no matter how many people are standing behind the bar.

The second layer is authority and role. Opening and closing usually require keyholder status and cash-handling training. A supervisory shift requires someone who can actually make the decisions the shift makes. Sites with ratio requirements, whether legal ratios in care settings or internal ones such as one experienced person per shift, break quietly when a swap replaces the wrong person with the wrong person.

The third layer is the one people forget: a swap between two people at different pay rates does not change what either of them earns. Each employee is paid their own rate for the hours they actually work. If a higher-rated employee picks up a lower-rated colleague's shift, your labour cost for that shift rises, and if the trade also crosses forty hours you are paying the higher rate at time and a half.

The cheapest fix is a swap-eligibility matrix, which is one line per role listing what a person must hold to work it. Keep it next to the schedule so the approver checks it in one glance rather than from memory. Cross-training widens the matrix over time and is the only durable answer to a small pool of eligible people.

Minors, and the Hours a Swap Cannot Break

If either employee in a trade is under eighteen, the age rules override the agreement completely. Federal child labor rules cap hours and set a time-of-day window for fourteen and fifteen year olds, and neither the minor nor a parent can consent past them, which makes age a hard gate rather than a judgement call.

The federal limits for fourteen and fifteen year olds are specific: no more than three hours on a school day and eighteen hours in a school week, no more than eight hours on a non-school day and forty hours in a non-school week, and work confined to the period between 7 a.m. and 7 p.m., extended to 9 p.m. from 1 June through Labor Day (29 CFR 570.35). A Saturday-for-Thursday trade that looks harmless can breach the daily cap on its own.

Sixteen and seventeen year olds have no federal hour cap, which misleads a lot of employers, because the hazardous occupation orders still bar them from specific tasks and many states impose their own nightly cutoffs, daily maximums, and school-night rules. State law is frequently stricter than federal law here, and where they differ the stricter one governs. The state picture is in the guides to child labor laws and hours a 16 year old can work.

Two operational rules keep this manageable. Flag minor status in the schedule itself, so the approver sees it without remembering birthdays, and require that any swap involving a minor be approved by a named person rather than whoever is on shift. Penalties are assessed per minor and per violation, so a single misjudged trade on a busy Saturday can multiply quickly.

Work permits and school-hours certificates add a second layer in many states, and they are date-sensitive in a way that trips swaps specifically: a permit valid for one schedule pattern may not cover the shift the minor picks up. Keep the permit with the employee record and check the expiry when you check the hours.

Fairness, and the People Who Absorb Every Bad Shift

Left alone, swap systems concentrate. A small group ends up absorbing most of the weekends, closes, and holidays, and that group is usually the newest, the most junior, or the least comfortable saying no. The pattern is invisible until somebody counts, and it is one of the quietest causes of turnover in scheduled businesses.

The mechanism is simple. Requests travel through the same two or three phone numbers, because those are the people who always say yes, and each individual trade is voluntary and reasonable. Fifteen reasonable trades later, one person has worked six of the last eight Saturdays and is looking at other jobs, while the schedule you published looks perfectly balanced.

Three controls fix most of it without adding process. Broadcast, so a swap request goes to everyone eligible for that role rather than to a private list, which both spreads the load and improves your odds of finding a taker. Cap, so no individual absorbs more than a set number of traded shifts in a month, which forces requests outward. And count, so you review the tally once a quarter and see the concentration before the person resigns.

The counting is the part people skip and the part that pays. Pull one quarter of approved swaps and tabulate two columns: shifts given up by each person and shifts absorbed by each person. In most teams the top absorber carries three or four times the median. A spreadsheet and a recurring calendar entry are enough, and the numbers usually make the conversation for you.

There is a compliance edge here too. If the same people are always the ones released from undesirable shifts, and that correlates with a protected characteristic, an even-handed policy applied unevenly starts to look like something else. Consistent, recorded decisions are the defence.

Getting the Swap Into the Timekeeping Record

An approved swap that never reaches the schedule and the time record creates two errors in the same pay period: the person who worked looks absent, and the person who did not work appears to have hours nobody clocked. Pay follows the record, so the record has to be reconciled before the period closes.

Federal wage and hour rules require accurate daily and weekly hours for every non-exempt employee, and the schedule you published is not that record. What matters is the hours actually worked, which is why a swap is only finished when the timekeeping side agrees with what happened on the floor.

The attendance consequence is the one that damages trust fastest. Under a point system, an unrecorded swap hands a point to the employee who did everything right and asked permission first. It takes one occurrence of that to teach a team that using the official process is riskier than trading privately, and undoing that lesson takes months.

Three habits close the loop. Update the posted schedule at the moment of approval rather than at the end of the week. Reconcile approved swaps against clock records before you run pay, treating any mismatch as an error to investigate. And keep the approval thread with the schedule record, so a year later you can still show who asked, who agreed, and who signed off.

The Policy Language Itself

A shift swap policy needs seven clauses and fits on a single page. Anything longer stops being read, and the point of writing it down is that a supervisor can apply it consistently at seven in the morning without calling anyone.

Write the payment rule in as well, in one sentence: each employee is paid their own rate for the hours they actually work, and any overtime created by an approved trade is paid at the required premium. It reads as obvious and it heads off the two conversations that otherwise happen every few months.

Put the policy in the employee handbook and walk through it during onboarding, not during the first crisis. Pair it with the request channel people already use, because a policy that requires a form nobody opens produces exactly the private text messages it was written to prevent.

Frequently Asked Questions