Every stat holiday puts a calculation decision on your plate. For most Canadian SMBs, that decision gets made the same way every time: copy what was done last time, assume it's correct, move on. That works fine until it doesn't, and when it stops working, you typically find out through a complaint, not a routine audit.
The problem is not that stat holiday pay is complicated. It isn't. The problem is that Canada has four different formulas, and they don't all produce the same number. If you're running the Ontario formula on your Alberta team, or the Alberta formula on your BC employees, you're getting the wrong answer. Possibly every single stat, every single year.
This guide walks through the exact calculation for each formula family, with worked examples, so you can run it confidently, or verify what your payroll software is doing.
The four formula families (and which provinces use each)
Before doing any math, you need to know which formula applies to each employee based on their province of work, not where your business is registered, and not where you're headquartered.
| Formula | Provinces |
|---|---|
| 1/20 of wages in prior 4 weeks | Ontario, Quebec, Federal (Canada Labour Code), Yukon |
| 5% of wages in prior 4 weeks | Alberta, Saskatchewan, Manitoba |
| Regular day's pay | New Brunswick, Nova Scotia, PEI, Newfoundland & Labrador, NWT, Nunavut |
| Wages in prior 30 days ÷ days worked | British Columbia (only) |
For employees covered by the federal Canada Labour Code (transportation, banking, telecommunications, federal Crown corporations), use the Ontario/Quebec 1/20 formula regardless of the province they work in.
Formula 1: Ontario, Quebec, Federal, Yukon, the 1/20 rule
The formula: Add up the employee's total wages earned in the four complete workweeks before the week of the holiday. Divide by 20.
Stat pay = Total wages in the 4 workweeks before the holiday week ÷ 20
Ontario nuance: Include any vacation pay that was paid during those four workweeks in the total wages.
Quebec nuance: Vacation pay is excluded. Use regular wages only.
Worked example, Ontario, weekly pay cycle
Sarah is an Ontario employee on a weekly pay cycle. The stat holiday is Victoria Day, Monday May 19, 2026.
The four workweeks before the holiday week (ending May 18) are:
- Week ending April 27: $1,200
- Week ending May 4: $1,100
- Week ending May 11: $950 (took a sick day)
- Week ending May 18: $1,200
Total = $4,450. Vacation pay of $178 was paid on May 4. Ontario total (wages + vacation pay) = $4,628.
Stat pay = $4,628 ÷ 20 = $231.40
That's what Sarah is owed for Victoria Day if she doesn't work it. If she does work it, she also earns 1.5× her regular rate for hours worked on top of this amount.
Formula 2: Alberta, Saskatchewan, Manitoba, the 5% rule
The formula: Add up the employee's total wages earned in the 4 workweeks before the stat holiday (not necessarily "before the holiday week", just the 28 calendar days before the holiday). Take 5% of that total.
Stat pay = Total wages in the 4 weeks before the holiday × 5%
The 5% rule and the 1/20 rule look similar, and for a full-time employee working consistent hours they produce nearly identical results ($4,500 × 5% = $225 vs. $4,500 ÷ 20 = $225). The divergence appears with part-timers, employees on variable schedules, and anyone who took unpaid time in that window.
Worked example, Alberta, bi-weekly pay cycle
Marcus is an Alberta employee on a bi-weekly cycle. The stat holiday is Alberta Family Day, Monday February 16, 2026.
The 4 weeks before February 16:
- Jan 19–Feb 1 (bi-weekly pay period): $2,600
- Feb 2–Feb 15 (bi-weekly pay period): $2,200
Total wages = $4,800.
Stat pay = $4,800 × 5% = $240.00
Marcus doesn't work the stat. He gets $240.00 for Family Day. If he had worked the holiday, Alberta requires you to pay the $240.00 stat pay plus 1.5× his regular rate for hours actually worked.
Formula 3: New Brunswick, Nova Scotia, PEI, Newfoundland & Labrador, NWT, Nunavut, regular day's pay
The formula: Pay the employee what they would normally earn in a regular workday.
For salaried employees, this is straightforward:
Stat pay = Annual salary ÷ 260 working days
For hourly employees with consistent schedules, it's:
Stat pay = Normal daily hours × hourly rate
For hourly employees with irregular schedules, use the average daily hours over a meaningful recent window, typically the number of hours they'd typically work on that day of the week.
Worked example, Nova Scotia, hourly employee
Lin works Monday through Friday at $22/hr, 8 hours per day. The stat holiday is Nova Scotia Heritage Day (the third Monday in February).
Lin's regular day's pay = 8 hours × $22.00 = $176.00.
That's the stat pay. Simple.
Nova Scotia specific note: Nova Scotia only has 6 statutory holidays per year, the fewest in Canada. Make sure you're not paying for days that aren't actually Nova Scotia stats, and don't miss the ones that are.
Newfoundland specific note: Newfoundland has 6 paid public holidays under the Labour Standards Act, tied with Nova Scotia for the fewest in Canada. If an employee works the stat, the premium rate is 2×, not 1.5× like most other provinces.
Formula 4: British Columbia, the 30-day average
BC uses a formula no other province uses. It is more precise than the others but also easier to apply incorrectly.
The formula:
Stat pay = Total wages earned in the 30 calendar days before the holiday ÷ Number of days worked in those 30 days
"Days worked" means days the employee actually performed work, not paid days off, not sick days, not the stat holiday itself.
Worked example, BC, salaried part-time employee
Priya works 3 days a week in BC at a salary equivalent to $900/week. The stat holiday is BC Day, Monday August 3, 2026.
In the 30 days before August 3 (July 4 to August 2), Priya worked 13 days and earned $3,600 in wages.
Stat pay = $3,600 ÷ 13 = $276.92
Compare that to the 1/20 rule: $3,600 ÷ 20 = $180.00. The BC formula produces $96.92 more for this employee. Using the Ontario formula on a BC employee systematically underpays part-timers. That gap compounds across every stat in the year, BC has 10 of them.
