The Employer-Designated Workweek: Why the Day Your Pay Week Starts On Decides Your Overtime

Ask most people when their "week" starts for pay purposes and they'll say Sunday, or Monday, or "whenever the pay period starts." None of those are necessarily right, and the gap between what people assume and what's actually on file with payroll is where a real amount of confusion about overtime comes from.

Under the FLSA, overtime isn't computed over a pay period, and it isn't computed over the calendar week either. It's computed over a workweek β€” a specific, employer-defined 168-hour block that your employer chose, wrote down, and is supposed to keep fixed. Where that block starts is not a detail. On a schedule with shifts that cross midnight or a rotation that doesn't divide evenly by seven, it decides which hours land in which week β€” and therefore whether hours 41 and up exist at all.

What a "workweek" actually is

The definition comes straight from the regulations, and it's worth reading in full because almost nothing in it matches the calendar-week assumption:

"An employee's workweek is a fixed and regularly recurring period of 168 hoursβ€”seven consecutive 24-hour periods. It need not coincide with the calendar week but may begin on any day and at any hour of the day." (29 CFR Β§ 778.105)

Three things in that sentence matter more than they look like they should. It's fixed β€” not something that floats to match whatever schedule happened to run that period. It's regularly recurring β€” the same 168 hours, week after week, not redrawn to fit the roster after the fact. And it can start "on any day and at any hour of the day" β€” Wednesday at 6:00 p.m. is exactly as valid a workweek boundary as Sunday at midnight.

The same section adds that the employer can set one workweek for an entire site or different workweeks for different groups of employees, and that once it's set, "it remains fixed regardless of the schedule of hours worked."

And overtime is computed against that single block, not smoothed across two of them. The companion rule is blunt about it:

"The Act takes a single workweek as its standard and does not permit averaging of hours over 2 or more weeks. Thus, if an employee works 30 hours one week and 50 hours the next, he must receive overtime compensation for the overtime hours worked beyond the applicable maximum in the second week, even though the average number of hours worked in the 2 weeks is 40." (29 CFR Β§ 778.104)

That's what "each workweek stands alone" means in practice. A slow week never buys back a heavy one, and the underlying overtime requirement β€” 29 U.S.C. Β§ 207(a)(1) β€” is written the same way, requiring time and a half for hours "in excess of" forty "in any workweek," full stop, no reference to pay periods, calendar weeks, or averages.

None of this means your pay period has to be weekly. Plenty of employers run biweekly or semi-monthly pay periods for cutting checks, while the underlying overtime math is still done one 168-hour workweek at a time and the results get added into whatever period you're actually paid on. The pay period is an accounting convenience. The workweek is the unit the law actually measures.

Why the start day changes the number, not just the label

Because the workweek is a fixed 168-hour block and overtime doesn't average across blocks, where the block starts determines which side of the boundary each hour falls on β€” and that can change whether overtime exists at all, without changing a single hour anyone actually worked.

Take a schedule that doesn't divide evenly by seven β€” a lot of rotations don't. Depending on where the employer's workweek boundary happens to land inside that rotation, a stretch of shifts can get sliced into two workweeks that both stay under forty, or into one heavy workweek that clears forty and one light one that doesn't. Same roster, same hours, different boundary, different overtime total. Nobody worked more; the accounting unit just landed differently.

This is also exactly the mechanism behind the 9/80 compressed schedule's extra day off β€” the employer deliberately moves the workweek boundary to the midpoint of the short Friday specifically so two otherwise-uneven weeks (44 and 36 hours) both land at a clean 40. That's covered in more detail in the 4/10 and 9/80 explainer; the workweek boundary is the same lever, just used deliberately there instead of incidentally.

A shift that crosses midnight, split by the boundary

Here's the case that trips people up most often: a night shift that starts before the workweek boundary and ends after it.

Say the designated workweek runs Saturday 00:00 through the following Friday 23:59, and a shift runs Saturday 19:00 to Sunday 07:00 β€” a normal overnight. That shift doesn't belong to one workweek. It's split:

Portion of the shift Clock time Which workweek it counts toward
First segment Saturday 19:00 – Saturday 23:59 (5 hrs) The workweek that just ended
Second segment Sunday 00:00 – Sunday 07:00 (7 hrs) The workweek that just began

Twelve hours worked, on one continuous shift, credited to two different 168-hour blocks β€” because the block, not the shift, is what the law measures. Move the designated boundary to a different day or hour, and the same shift could land entirely inside one workweek instead of being split at all. Nothing about the shift changed. Only where the accounting line falls did.

This is why "how many hours did I work this week" and "how many hours did I work in my designated workweek" aren't always the same question, especially for anyone whose shifts routinely straddle midnight.

How to find your designated workweek

It's a specific fact your employer has on file β€” not something to reverse-engineer from a gut sense of "the week." A few places it actually shows up:

If none of those turn up an explicit answer, that's itself worth asking about β€” the regulation requires the employer to have designated one, even if it happens to line up with the calendar week.

What it takes to change it

An employer isn't locked into a workweek forever, but a change has to clear a specific bar, and it's a narrow one:

"The beginning of the workweek may be changed if the change is intended to be permanent and is not designed to evade the overtime requirements of the Act." (29 CFR Β§ 778.105)

Two conditions, both required:

  1. Intended to be permanent β€” a real, ongoing change to the fixed schedule, not something toggled for a pay period and switched back.
  2. Not designed to evade overtime β€” the change can't exist to manufacture a boundary that happens to split a heavy stretch of hours into two workweeks that each stay under forty.

A legitimate reason to move it is common enough β€” a shift in business hours, a new payroll system, aligning multiple sites onto one schedule. What the rule is actually policing is a boundary that moves specifically to cut a real 45-hour stretch into two paper 22.5-hour ones, or that moves back and forth to chase whatever produces the least overtime that period. The regulation also notes that a period straddling an actual change in workweek has its own transition computation, so a shift in the boundary isn't self-executing β€” there's a documented method for paying out the overlap correctly.

Checklist: what to actually verify

Worth knowing if you're comparing notes with someone else

Two shift workers can run the exact same roster β€” same shift lengths, same days on, same rotation β€” and land on different overtime totals for the same stretch, purely because their employers designated different workweek start points. That's not a payroll error to chase down on its own; it's the regulation working as written. The number worth comparing isn't "how many hours did we each work this week" β€” it's where each of your employers actually drew the boundary.

One more thing worth flagging if you work for a multinational company or split time across borders: this whole framework β€” the fixed 168-hour block, the "each workweek stands alone" rule β€” is specific to the FLSA. Other countries compute overtime differently, sometimes over the calendar week only, sometimes averaged over a longer reference period. If you're not working under US federal law, the mechanism above doesn't automatically transfer, and it's worth checking what standard actually applies to you.

None of this is legal advice, and a real dispute over which hours landed in which workweek is a payroll or wage-and-hour question, not something to resolve from a blog post. The primary sources are linked above so you can read the actual clause rather than a summary of it β€” the next step, if the boundary and your rotation genuinely don't line up the way payroll says, is your employer's payroll or HR team, or a state labor agency.

Where the boundary actually falls on your rotation

A designated workweek is easy to reason about in the abstract and hard to track in practice, because it requires lining up a fixed clock-time boundary against a rotation that keeps moving underneath it. A shift that splits across the line one week won't necessarily split the same way the next, depending on how your pattern repeats.

That's the kind of thing Duty Pals is built to lay out automatically: set your rotation once, and see years of on and off days mapped out in advance, so a workweek boundary crossing your shift is something you can see coming rather than something you notice for the first time on a paycheck. Duty Pals is currently in pre-registration; you can sign up to be notified when it launches.