Overtime regulation

Overtime regulation requires employers to pay a premium, usually time and a half, for hours beyond a standard workday or workweek. It raises the price of long hours relative to the price of hiring another worker. Firms respond by cutting overtime and setting more workdays at exactly the standard length; the effect on the number of workers is ambiguous.

The rules

In Alberta, overtime is any time worked over 8 hours a day or 44 hours a week, whichever gives more overtime hours, and it must be paid at least 1.5 times the regular wage. A written overtime agreement can replace the pay with paid time off at one hour per overtime hour.¹ Federally regulated workplaces use standard hours of 8 a day and 40 a week, with overtime paid at 1.5 times the regular wage or taken as time off at 1.5 hours per overtime hour.² In the United States, the Fair Labor Standards Act of 1938 requires time and a half after 40 hours a week,³ and California also requires it after 8 hours in a day.⁴

Two prices of labour

A firm buys labour in two units: workers and hours per worker. Some costs are paid per hour, chiefly the wage. Others are paid per worker regardless of hours: recruiting, training, equipment and many benefits. Payroll taxes partly fall in this group, because employer EI premiums and CPP contributions stop once a worker’s earnings reach annual ceilings.⁵

Fixed costs per worker make longer hours the cheaper way to buy labour, up to the point where fatigue lowers output per hour. An overtime premium raises the price of hours past the standard while leaving the price of a worker unchanged. That changes relative input prices, and the long-run demand framework applies:

• Substitution effect: hours past the standard are now relatively expensive, so firms use fewer hours per worker and more workers. This is the “spread the work” case for overtime rules.
• Scale effect: labour costs more in total, so firms produce less and use less of both workers and hours.

Hours per worker fall for certain. Total hours fall. The number of workers rises only if the substitution effect outweighs the scale effect.

Hours per worker: a jump in the price of an hour

With a daily standard of 8 hours, the marginal cost of an hour is the straight-time wage up to 8 hours and the overtime rate after. A firm sets each worker’s day where the value of the marginal product of an extra hour, which falls with fatigue, meets this marginal cost.

Overtime premium

50%

Output price (p)

$4.00

Hours per day without the rule

10.0

Hours per day with the rule

8.0

Overtime hours cut per worker

2.0

Open in Graph Lab

Figure 1. The value of the marginal product of an extra hour per worker against the marginal cost of an hour, which jumps from the $20 straight-time wage to the overtime rate after 8 hours. With no rule the firm sets a 10-hour day; with a 50 percent premium the $30 overtime rate exceeds the value of every hour past 8, so the day is cut to exactly 8 hours. At a higher output price of $5.50 the firm still works overtime but cuts the day from 11.4 to 9.5 hours.

The jump in marginal cost produces three predictions:

1. Fewer overtime hours. Where the value of a ninth hour exceeds the straight-time wage but not the overtime rate, the hour is no longer worked.
2. Bunching at the standard. Any firm whose value of an hour at 8 lies between the two rates chooses exactly 8 hours, so 8-hour days become more common.
3. No change below the standard. Days that were already 8 hours or shorter face the same marginal cost as before.

Workers versus hours

When each worker carries a fixed daily cost, the cost of labour services depends on the length of the day. Short days spread the fixed cost over few hours; long days run into fatigue. The cheapest workday balances the two, and an overtime premium shortens it.

Overtime premium

50%

Fixed cost per worker per day (F)

$100

Cheapest workday without the rule

10.0 h

Cheapest workday with the rule

8.0 h

Extra workers needed for the same labour services

16%

Rise in cost per unit of labour services

0.6%

Open in Graph Lab

Figure 2. Cost per unit of labour services against the length of the workday, with a $20 wage, a fixed cost of $100 per worker per day and diminishing returns to longer days. Without the rule the cheapest day is 10 hours. A 50 percent premium makes the cheapest day 8 hours: the firm needs 16 percent more workers for the same labour services (substitution), and its cost per unit of labour services rises by 0.6 percent (the source of the scale effect).

Key idea

An overtime premium raises the price of hours relative to the price of workers. Hours per worker fall; whether employment rises depends on whether the substitution effect toward workers outweighs the scale effect of higher labour costs.

Wages can blunt the rule. If workers and firms care about total pay for a job with given hours, a firm can lower the straight-time wage so that pay for a long week is unchanged. Hours then need not fall. Trejo found that large expansions of US overtime coverage, brought about by legislative amendments and Supreme Court decisions, had no detectable effect on overtime hours once trends within industries were accounted for, consistent with straight-time wages adjusting to neutralize the premium.⁶

Evidence: California extends daily overtime to men

California required most women to receive time and a half after 8 hours in a day long before federal overtime rules existed. After the Civil Rights Act of 1964 the women-only rule was challenged as discriminatory. From 1974 legal battles left it largely unenforced for both sexes, and in 1980 the daily standard was extended to men.³ Course materials sometimes date the change to 1974; the analysis below compares 1973, when only women were covered, with 1985, when both were.

Hamermesh and Trejo used Current Population Survey data from those two years and the method of difference-in-differences: compare the change in an outcome for the group the rule affected (California men) with the change for a control group facing the same other shocks but not the rule. The control groups were men outside the West, where no daily overtime standard applied, and California women, who were covered in both years.

Share of workers whose day exceeds 8 hours	1973	1985	Change (points)
California men (treated)	17.12%	16.89%	$-0.23$
Men outside the West	20.13%	22.79%	$+2.66$
California women	3.96%	7.19%	$+3.23$

Long workdays became more common for both control groups but not for California men. Relative to men elsewhere, the share of California men working more than 8 hours a day fell by 2.9 percentage points, a 17 percent decline; relative to California women, by 3.5 points. Comparing both at once (a triple difference) gives a fall of 5.2 points, or 30 percent. The share working exactly 8 hours rose by 3.4 to 7.7 points relative to the controls, matching the predicted bunching, and men who still worked overtime worked about 0.28 fewer overtime hours a day. The authors estimate that the price elasticity of demand for overtime hours is at least $-0.7$, which they regard as an understatement, because the rule raised the price of overtime by less than 50 percent and workers may have wanted more overtime once it paid more.³

Average annual hours actually worked per worker

Selected OECD countries, all workers

• Canada
• United States

Figure 3. Average annual hours actually worked per worker in Canada and the United States, 1970 to 2025. Hours per worker are the margin overtime rules act on; the chart does not isolate the rules' effect.Source: OECD Data Explorer, Average annual hours actually worked per worker (OECD.ELS.SAE, DSD_HW@DF_AVG_ANN_HRS_WKD). Retrieved 2026-10-23.

Show the data and notes

Total hours actually worked in the year ÷ average number of people in employment (employees and self-employed). The OECD advises using these data to compare trends over time, not levels across countries, because national sources and methods differ. Rounded to whole hours, the OECD's published precision.

Period	Canada	United States
1970	1,925	1,949
1971	1,912	1,936
1972	1,904	1,933
1973	1,898	1,934
1974	1,891	1,903
1975	1,872	1,882
1976	1,862	1,884
1977	1,834	1,888
1978	1,836	1,879
1979	1,841	1,877
1980	1,827	1,859
1981	1,812	1,855
1982	1,789	1,847
1983	1,779	1,867
1984	1,789	1,884
1985	1,795	1,883
1986	1,797	1,875
1987	1,807	1,884
1988	1,808	1,884
1989	1,802	1,895
1990	1,797	1,878
1991	1,775	1,868
1992	1,772	1,867
1993	1,769	1,876
1994	1,778	1,887
1995	1,775	1,886
1996	1,790	1,887
1997	1,802	1,894
1998	1,797	1,892
1999	1,795	1,889
2000	1,789	1,880
2001	1,775	1,860
2002	1,757	1,854
2003	1,747	1,844
2004	1,757	1,845
2005	1,747	1,842
2006	1,745	1,847
2007	1,745	1,842
2008	1,742	1,830
2009	1,710	1,799
2010	1,715	1,810
2011	1,711	1,820
2012	1,721	1,827
2013	1,714	1,825
2014	1,709	1,830
2015	1,710	1,831
2016	1,705	1,822
2017	1,694	1,820
2018	1,704	1,826
2019	1,693	1,823
2020	1,658	1,800
2021	1,691	1,819
2022	1,693	1,810
2023	1,694	1,805
2024	1,697	1,796
2025	1,687	1,800

Worked example

Use the 1973 and 1985 shares of men working more than 8 hours a day.

1. Change for the treated group. California men: $16.89 - 17.12 = -0.23$ percentage points.

2. Change for the control group. Men outside the West: $22.79 - 20.13 = +2.66$ points. This is the change California men would have had without the rule, under the assumption that both groups would otherwise have followed the same trend.

3. Difference-in-differences. $-0.23 - 2.66 = -2.89$ points. As a share of the 1973 level, $2.89 / 17.12 = 0.17$: the rule cut the incidence of long days among California men by about 17 percent.

4. A second control. With California women as the control: $-0.23 - 3.23 = -3.46$ points.

5. Interpreting the elasticity. With an elasticity of demand for overtime hours of $-0.7$, a 50 percent rise in the price of an overtime hour cuts overtime hours by about $0.7 \times 50 = 35$ percent.

Exercise: the triple difference

Among women outside the West, the share with days over 8 hours went from 5.97 to 6.89 percent. The difference-in-differences for women (California minus elsewhere) is $(7.19 - 3.96) - (6.89 - 5.97) = 3.23 - 0.92 = 2.31$ points; the rule did not change for women, so this measures California-specific shocks. Subtracting it from the men’s estimate gives $-2.89 - 2.31 = -5.20$ points, the triple difference.

Common mistakes

Traps on exams

• Assuming overtime rules must create jobs. The substitution effect favours more workers, but the scale effect reduces all labour, and straight-time wages may adjust. The employment effect is ambiguous.
• Saying the rule raises the cost of every hour. It raises the marginal cost of hours past the standard only, which is why days bunch at exactly 8 hours.
• Skipping the control group. California men’s share barely changed between 1973 and 1985; the effect appears only against the rise in long days elsewhere.
• Mixing percentage points and percent. A fall from 17.1 to 14.2 percent is 2.9 percentage points, or 17 percent.

Check yourself

1.

A province introduces an overtime premium for hours beyond 8 a day. According to the labour demand framework, what happens to hours per worker and to the number of workers?

1Hours per worker fall; the number of workers may rise or fall2Hours per worker fall and the number of workers rises, because the work is spread out3Both hours per worker and the number of workers fall, because labour costs rise4Neither changes, because the premium is a transfer from firms to workers

Check answer

2.

Between 1973 and 1985, the share of California men working more than 8 hours a day went from 17.12 to 16.89 percent. Among men outside the West, where no daily overtime rule applied, it went from 20.13 to 22.79 percent. What is the difference-in-differences estimate of the effect of California's daily overtime rule, in percentage points?

Your answer (percentage points)

Check answer

3.

Which development would most threaten the difference-in-differences estimate of the effect of California's daily overtime rule on men's long workdays?

1Industries that use long shifts shrank in California but not elsewhere between 1973 and 19852Long workdays became more common everywhere in the United States3California men worked more days per week in both years4The federal 40-hour overtime rule applied to both groups throughout

Check answer

Practise more questions on this topic

Sources

1. Government of Alberta, Overtime hours and overtime pay. ↩

2. Government of Canada, Work hours: federal labour standards. ↩

3. Hamermesh, D. S. and Trejo, S. J. (2000). The demand for hours of labor: Direct evidence from California. Review of Economics and Statistics, 82(1), 38 to 47. Figures from NBER working paper 5973 (1997), Table 1 and section IV. ↩ ↩² ↩³

4. California Department of Industrial Relations, Overtime: frequently asked questions. ↩

5. Canada Revenue Agency, EI premium rates and maximums and CPP contribution rates, maximums and exemptions. ↩

6. Trejo, S. J. (2003). Does the statutory overtime premium discourage long workweeks? Industrial and Labor Relations Review, 56(3), 530 to 551. IZA Discussion Paper 373. ↩