The Shift That Crosses Midnight
A night shift belongs to two dates and every system has to pick one. The choice affects daily thresholds, weekly totals, premiums and the clock change.
One night shift, four questions
| What happens | Required by | Paid? |
|---|---|---|
| Which date does it belong to | Configuration | Contested |
| Which week does it count in | Configuration | Contested |
| Does the daily threshold reset at midnight | Configuration | Contested |
| Does the premium apply to both halves | Agreement | Paid |
Three of the four are settings rather than agreements. They were chosen at implementation and they determine a night worker's pay every week.
A shift from 22:00 to 06:00 happens on two dates. Every calculation that uses a date has to decide which, and there is no natural answer.
The practical lesson in “The Shift That Crosses Midnight” is that a timesheet only becomes reliable through a process people can operate consistently. For teams exploring time tracking with screenshots, a practical route to time tracking with screenshots can add time and project context, provided collection is proportionate, access is limited and every consequential inference receives human review.
The decision is a configuration setting. It affects the daily hours total, the weekly total, which week the shift falls in, whether a daily threshold resets mid-shift, and how the shift is reported. For night workers it is one of the largest determinants of pay and almost none of them know it exists.
For a separate benchmark relevant to “The Shift That Crosses Midnight”, consult the WHO mental-health-at-work resources. Use it to test record quality, approvals, retention, employee rights and exception handling against the real workflow rather than treating a software report as self-explanatory evidence.
The two conventions
Attribute to the start date. The whole shift counts on the day it began. Most common and generally the more intuitive for shift workers, who think of Monday night as Monday.
Split at midnight. Hours before midnight on one date, after on the next. Produces results that match a calendar and that night workers find strange, because a single shift appears as two part-days.
Either is defensible. What is not defensible is not knowing which one is in use, which is the normal state.
What it changes
Daily thresholds. With a split, a shift of eight hours appears as two hours and six, so a daily overtime threshold of eight is never crossed by a night shift however long it runs.
Weekly boundaries. A shift starting Sunday 22:00 falls in one week under one convention and spans two under the other. For a worker close to a weekly threshold this decides whether they cross it.
Consecutive-day counts. Rest-day and consecutive-shift rules behave differently, and a night worker can appear to have worked seven days when they worked five shifts, or the reverse.
Reporting. Attendance by day, absence by day, and productivity by day all shift, which matters for comparisons between day and night operations.
The clock change
Twice a year the night shift is seven hours or nine. Most systems handle this correctly and some do not, and the ones that do not get it wrong in the same direction every year.
Check both affected shifts by hand, annually. It is two records and it catches an error that otherwise repeats indefinitely.
Where the agreement says what happens — paid for the rostered hours regardless, or paid for the actual hours — the configuration should match it, and frequently does not.
Night-work rules specifically
Many jurisdictions define a night worker by reference to hours worked in a defined night period, and attach consequences: tighter maximum hours, health assessment entitlements, specific record-keeping.
Whether somebody qualifies depends on how their hours fall inside that period, which depends on the midnight convention. A system that attributes everything to the start date may not be able to compute hours within the night window at all, which means the organisation cannot tell who its night workers are.
That is a compliance gap rather than a pay one, and it is common.
Finding out what yours does
Take one real night shift and follow it through: what date does it appear on, how many hours on each date, did it cross a daily threshold, which week did it land in.
Then take a shift that starts on a Sunday night and do the same. The Sunday case exposes the weekly boundary behaviour, which is where the surprises are.
What to write down
The convention, the reason for it, how the weekly boundary works, whether daily thresholds reset at midnight, and how the clock change is handled.
Five lines, held with the pay-rule documentation. It is the thing nobody can answer when a night worker asks why their week looks the way it does, and the question is entirely reasonable.
What the person sees on the payslip
Whichever convention is used, the night worker's week on paper will not look like the week they experienced, and no explanation printed on the payslip fixes that on its own.
What helps is a view of the week as recorded, with each shift shown against the date it was assigned to and a line saying which convention applies. It takes a report rather than a change to the rules, it removes most of the queries, and it means that when the week genuinely is wrong the person can see it in the first place rather than only the total.