Salaried Overtime Requirements: Employers Will Offset Them with Lower Pay

Report Jobs and Labor

Salaried Overtime Requirements: Employers Will Offset Them with Lower Pay

July 2, 2015 13 min read Download Report
James Sherk
James Sherk
Research Fellow, Labor Economics
As research fellow in labor economics at The Heritage Foundation, James Sherk researched ways to promote competition and mobility.

The Obama Administration has announced plans to require overtime pay for salaried employees who earn less than $50,440 a year. Economic research shows that employers will offset new overtime costs by lowering base salaries. As a result, these regulations will have little effect on total weekly earnings or hours worked. They will require employers to rigidly monitor salaried employees’ hours. This would proscribe the flexible working arrangements that many salaried employees value. These regulations will limit workplace flexibility without improving pay.

New Overtime Regulations Proposed

The U.S. Department of Labor has announced plans to significantly alter its overtime regulations. The Fair Labor Standards Act (FLSA) requires most employers to pay their employees one and a half times their usual pay for time worked above 40 hours a week. The FLSA also exempts certain employees from overtime requirements—principally executives, administrative and professional employees, as well as salesmen who work outside the office. The FLSA charges the Department of Labor with defining these categories through regulation.

Under the existing regulations, employers must pay overtime to all hourly workers. Employers must also pay overtime to salaried employees who either earn less than a certain amount (the salary test) or do not have sufficiently advanced job duties (the duties test). Employers must track the hours of any salaried employee eligible for overtime.

The Obama Administration has proposed significantly increasing the salary test from $455 a week ($23,660 a year) to $970 a week ($50,440 a year). Employers would have to track the hours and pay overtime to any salaried employee making less than this amount—no matter how advanced their job duties.

Intended to Raise Pay and Redistribute Work

Liberal groups have extensively lobbied the White House to implement these changes.[1] They argue that employers “misclassify” millions of salaried employees as exempt from overtime pay. They contend that this misclassification allows employers to compel these employees to work long hours without additional pay. Increasing the salary threshold to $50,440 a year would force employers to pay salaried employees with moderate incomes overtime no matter their job duties. The Administration and its allies argue that mandatory overtime rates will raise the earnings of 5 million salaried employees.

The Administration also hopes that these changes will reduce unemployment and underemployment by redistributing work. They intend the overtime requirements to encourage employers to hire new workers or give part-time employees longer hours instead of having existing employees work longer hours.

Economic research, however, shows that expanding overtime coverage does not accomplish these goals. Employers largely respond to new overtime requirements by cutting base pay—leaving total hours and earnings little changed.

Two Models of Employment. Economists have developed two models to explain how employers respond to changes in overtime laws: the “fixed-wage” and “fixed-job” models.[2] The fixed-wage model holds that employers treat hourly wage rates as a given determined by outside market forces.[3] Mandatory overtime makes hours worked above 40 hours a week more expensive. This higher cost encourages employers to substitute cheaper alternatives for overtime labor—either hiring new employees at regular hourly rates or making capital investments that reduce the need for labor. Employees who work overtime after these adjustments enjoy higher pay. The Administration has based its overtime regulations on this fixed-wage model.

However, even economists who subscribe to the fixed-wage model find that expanding overtime probably does not increase employment in the long term. They estimate that employers respond with additional labor-saving capital investments that counteract the incentives to hire new workers. Daniel Hamermesh, author of Labor Demand, the seminal book on the topic, warns of

the importance of considering capital-labor substitution in evaluating the impact of changing the rate of premium pay.… Using the best available estimates of the extent of substitution along the various margins of labor demand, it seems clear that higher overtime pay will not expand employment, unless labor supply is far more elastic than the huge array of estimates suggests.[4]

By contrast, the fixed-job model posits that employers and employees negotiate an employment contract covering both wages and hours—hourly wages depend partly on how many hours the employee works, including overtime.[5] In this model, employers and employees care only about the total hours worked and total pay for that work, not the wages for particular hours. This model predicts that employers can pay workers less than the going wage for regular hours if the employee expects compensating overtime pay.

Under the fixed-job model, overtime requirements do not affect pay or hours. Employers and employees respond to the higher overtime costs by reducing regular wages, leaving total earnings unchanged. Unless workers earn close to the minimum wage—in which case their employers cannot offset the overtime with lower wages—overtime requirements have little effect on hours or earnings.[6]

Economic Research on Overtime and Earnings

Empirical research generally supports the fixed-job model rather than the fixed-wage model. The appendix summarizes recent research on overtime rates, hours worked, and employee earnings. While some studies find that overtime laws reduce average workweeks, most find that they have little effect on average hours worked. Moreover, most studies find that employers offset all or almost all of the cost of overtime premiums with lower base wages. One recent study found that employees paid for 80 percent of the cost of overtime coverage through lower regular wages.[7]

Even left-leaning economists recognize this fact. Jared Bernstein, the former chief economist for Vice President Joe Biden, and Ross Eisenbrey, vice president of the liberal Economic Policy Institute, wrote recently in support of increasing the overtime salary threshold.[8] Nonetheless, they recognize the Administration’s requirement will not increase average pay. They write that the “costs of increased coverage would ultimately be borne by workers as employers set base wages taking expected overtime pay into account.”[9] Indeed, Bernstein and Eisenbrey criticize analysts who argue that expanding overtime raises hiring costs, concluding that

this line of argument erroneously assumes that the incidence falls on the employer, not the worker. Labor economists consistently assume otherwise—that the incidence falls on the worker—which in this case means that the wage offer reflects expected overtime hours, as shown in footnote two. As such, there is no change [in labor costs] at the margin from expanding coverage, at least once the pay of newly covered, existing workers is allowed to adjust.[10]

Bernstein and Eisenbrey’s footnote 2 shows the formula that employers will use to calculate (or reduce) base wages, taking expected overtime rates into account.[11] Economists of all political persuasions agree that expanding overtime coverage has little effect on overall earnings.[12]

Recent experience also demonstrates this fact. A court decision in Japan extended overtime rates to “name only” managers at McDonald’s restaurants who do work similar to that of hourly employees. McDonald’s complied with the ruling, but reduced pay by an offsetting amount.[13] Similarly, IBM recently reclassified 7,000 salaried and technical-support workers as eligible for overtime as part of a lawsuit settlement. It also cut their base pay by 15 percent, leaving their total earnings unaffected.[14] Extending overtime to more salaried employees will have little effect on either their hours or earnings.

Reducing Workplace Flexibility

Extending overtime coverage would effectively turn millions of professional salaried employees into hourly workers. The Fair Labor Standards Act requires employers to pay covered employees time and a half for all hours worked in excess of 40 a week. Salaried employees cannot waive this requirement. Consequently, the FLSA forbids paying covered employees for completing a job, forcing them instead to pay on the basis of hours worked. Hourly or salaried, employers must carefully monitor covered employees’ hours and pay overtime.

This restricts the flexible work arrangements on which professional employees frequently rely. Many salaried employees can take time off as needed, provided they get the job done. For example, they could take their child to the doctor’s office during work hours, but their employer would expect them to finish the work later—such as at home in the evening. Similarly, many Americans work from home. Surveys estimate that 3.3 million employees primarily telecommute, while between 16 million and 25 million more telecommute at least once a month.[15]

However, the requirement to track hours forces employers to restrict flexible schedules and teleworking. Logging hours worked on a job site is simple—tracking hours worked at home is not. Each time overtime-eligible employees respond to a work e-mail, take a work phone call, or do any other work from home, their employer must track and pay them for it. If they do not, they risk getting sued. Trial lawyers filed over 8,000 FLSA lawsuits in 2013, many of them for employers who did not compensate overtime-eligible employees for work done remotely.[16]

In order to avoid lawsuits, many employers deny flexible work arrangements to overtime-eligible employees. Virtually all employers who permit remote work and flexible work arrangements allow overtime-exempt employees to use them. Only about half allow workers covered by overtime regulations to do so.[17] As the head of human resources for Pitney Bowes explained to reporters, the company turned down requests from overtime-eligible staff to work from home because: “You just don’t take the [legal] risk.”[18]

Restricting workplace flexibility will hurt salaried employees personally and professionally. Telecommuting and flexible work arrangements help employees balance their work and family responsibilities—increasingly important as two-income families become the norm, not the exception. Surveys show that such flexible work arrangements boost employee job satisfaction.[19] Many employers report that—used properly—they boost employee performance while reducing turnover and unplanned absences.[20] Covering millions of salaried professional employees under overtime regulations will deprive them of workplace flexibility without raising their pay.


The U.S. Department of Labor’s proposal to raise the overtime exemption threshold to $50,440 a year will have unintended consequences. While the White House wants to raise workers’ pay, academic and even liberal economists conclude that employers will respond with offsetting reductions in base wages. These regulations will have little effect on salaried employees’ total pay or hours worked.

Instead, they will effectively convert millions of salaried professional employees into hourly workers required to clock their time. Millions of workers will find that their employers can no longer pay them for a job completed instead of hours worked. While this makes little difference for employees with a fixed-job site, it will hurt those capable of working remotely. It will severely limit their use of the flexible work arrangements and telecommuting options that many rely on to balance their work and family lives. Expanding overtime regulations to more salaried employees will hurt the workers the White House wants to help.

—James Sherk is Research Fellow in Labor Economics in the Center for Data Analysis, of the Institute for Economic Freedom and Opportunity, at The Heritage Foundation. The author thanks Center for Data Analysis intern J. J. Deveney for his valuable contributions to this report.


Empirical Research on the Effects of Overtime Requirements

Daniel Hamermesh and Stephen Trejo (2000) found support for the fixed-wage model.[21] They examined how workweeks changed in California when the legislature required employers to pay men overtime for working more than eight hours a day—not just 40 hours a week—in 1980. Previously, the legislature had required employers to pay daily overtime to female employees. Hamermesh and Trejo found that the proportion of men in California working more than eight-hour days dropped by a quarter, and male employment rates increased relative to men in other states. However, almost all of this change came from among men already receiving weekly overtime under federal law. Daily overtime did not raise the cost of hiring these workers, so it should not have affected their hours. This study may have simply picked up pre-existing trends instead of an actual causal effect of the legislation.[22]

Dora Costa (2000) compared changes in hours in wholesale and retail trade—the former getting covered by the FLSA in 1938 and the latter in 1961.[23] In support of the fixed-wage model she found that overtime workweeks fell measurably in wholesale relative to retail in the 1940s. However, work hours fell the most for low-wage employees—those for whom employers would have the least room to offset the new requirements with lower wages. Similarly, Costa found much larger reductions in overtime hours in Southern states—which at the time had much lower wages than the rest of the country. Both these findings support the fixed-job model.

Jane Friesen (2002) examined Canadian data and found weak evidence consistent with the fixed-wage model.[24] She compared the distribution of weekly hours across Canadian provinces—each of which has the authority to set its own overtime threshold. She found that lower overtime thresholds correlate with lower average hours, although most of her estimates lack statistical significance. She also found that lower standard hours correlate with greater instances of moonlighting. In contrast to almost all other studies, Friesen also found that greater overtime coverage correlates with higher base-wage rates.

David Bell and Robert Hart (2003) examined data from the United Kingdom and found strong support for the fixed-job model.[25] The U.K. does not legally require firms to pay premium rates for overtime hours. Nonetheless, Bell and Hart found that the average British firm pays a 28 percent overtime premium. They further found that companies that pay higher-than-average overtime premiums pay lower base wages and vice versa—exactly what the fixed-job model predicts. Bell and Hart also found that average hourly earnings remain essentially constant irrespective of how many overtime hours employees work—again in line with the fixed-job model that predicts that employers and employees negotiate wages and hours jointly.

John Johnson (2003) analyzed the effects of a Supreme Court decision requiring most state and local government employers to pay overtime under the FLSA.[26] In support of the fixed-job model he found that the ruling had virtually no effect on the amount of overtime worked. However, far more government employees belong to unions than do private-sector workers.[27] Unions expressly bargain over labor contracts that cover both pay and hours—the fixed-job model most clearly applies to them. Johnson’s findings do not necessarily extrapolate to non-union private-sector workers.

Stephen Trejo (2003) examined overtime and hours using data on average overtime hours by industry between 1970 and 1989.[28] During these years, Congress and the Supreme Court successively expanded the FLSA to cover more industries. Unlike his earlier research on California, this research supported the fixed-job model. Trejo’s initial estimates found that employers reduce overtime hours worked by about one-eighth. Controlling for pre-existing trends in hours worked reverses the picture—extending overtime coverage had no statistically significant effect on average work hours. Trejo concluded that his “finding is consistent with a model of labor market equilibrium in which straight-time hourly wages adjust to neutralize the statutory overtime premium.”

Mikal Skuterud (2007) examined the effect of a Quebec law reducing the overtime threshold for non-union hourly workers from 44 hours to 40 hours a week.[29] He found evidence consistent with both the fixed-wage and fixed-job models. Consistent with the fixed-wage model, he found that the change reduced the overtime hours of less-skilled men by 20 percent, along with a slight increase in average pay for workers employed between 41 hours and 44 hours a week. Consistent with the fixed-job model, he found no effect on the overtime hours worked by women, no increase in employment rates (as would happen if employers hired more workers at regular hours), and lower base wages for employees working more than 44 hours a week.

Sachiko Kuroda and Isamu Yamamoto looked at the pay of “name-only managers” in Japan (2009).[30] These workers do essentially the same work as regular hourly employees but—by virtue of their managerial classification—do not qualify for mandatory overtime. In line with the fixed-job model, Kuroda and Yamamoto found that managers earn the same effective hourly rate as overtime-eligible employees. On an hourly basis, their higher pay equates to the overtime they would have earned as regular employees entitled to overtime rates. Regular workers reclassified as overtime-exempt name-only managers saw no change in their average work hours or hourly earnings.

Kuroda and Yamamoto (2012) subsequently examined name-only managers during a period of time that included the 2007–2010 economic downturn.[31] They found that even in the recessionary period, overtime-exempt employees earn the same average hourly pay as similar overtime-eligible employees—support for the fixed-job model. However, they found that in contrast to normal economic periods, overtime-exempt employees worked longer hours than overtime-eligible employees during the recession. This lends some support to the fixed-wage model during recessionary periods. They also report that in response to court decisions forcing McDonald’s to grant overtime eligibility to name-only managers, McDonald’s cut its base salaries by an offsetting amount.

Trejo (1991) examined data on household pay and earnings from the 1970s.[32] He concluded that employers offset roughly half the cost of overtime requirements through lower base wages.

More recent research by Anthony Barkume (2010) updated Trejo’s 1991 work using data from the 2000s.[33] Barkume corrected for two problems in Trejo’s original work: (1) Trejo ignored the value of benefits, which make up an increasingly large component of workers’ compensation, and (2) Trejo ignored the fact that employers would pay an overtime premium in the absence of a legal requirement to do so, as demonstrated by the British employers. Barkume found that ignoring this market premium caused Trejo to overstate how the FLSA raised overtime costs.

Barkume’s study found evidence supporting both models, but greater evidence for the fixed-job model. The fixed-wage model predicts that firms with costlier benefits will use more overtime. Substituting new employees for overtime hours costs these firms more—each new employee comes with an expensive benefit package, while overtime hours do not. Barkume indeed found evidence that this occurs, but he reports stronger evidence in favor of the fixed-job model. He found that employers offset 80 percent of new overtime costs with lower base wages. Further, he found less overtime use among employees near the minimum wage—workers whose pay could not fall.

[1] News release, “On President Obama’s Directive Expanding Overtime Pay,” National Employment Law Project, March 7, 2014, (accessed October 28, 2014).

[2] These are also known as the labor-demand and employment-contract models, respectively.

[3] Daniel S. Hamermesh, Labor Demand (Princeton, NJ: Princeton University Press, 1993), chapter 2.

[4] Ibid., p. 176.

[5] H. Gregg Lewis, “Employer Interest in Employee Hours of Work,” unpublished paper, Duke University, 1969, and Yoram Barzel, “The Determination of Daily Hours and Wages,” The Quarterly Journal of Economics, Vol. 87, No. 2 (May 1973), pp. 220–238.

[6] This adjustment may not occur immediately because employees resist base nominal wage cuts. Over a period of several years’ inflation, slower promotions and smaller raises allow employers to reduce regular real wages without cutting nominal pay. Thus, the fixed job model posits few long-term effects from an overtime expansion, after an initial adjustment period.

[7] Anthony Barkume, “The Structure of Labor Costs with Overtime Work in US Jobs,” Industrial and Labor Relations Review, Vol. 64, No. 1 (October 2010).

[8] Jared Bernstein and Ross Eisenbrey, “New Inflation-Adjusted Salary Test Would Bring Needed Clarity to FLSA Overtime Rules,” Economic Policy Institute Report, March 13, 2014.

[9] Ibid., p. 2.

[10] Ibid., p. 13.

[11] Their footnote 2 reads: “If w equals the hourly wage the employer plans to pay for a given worker and w_1 is the base wage offer, including OT costs, TOT equals total weekly hours, and OT equals overtime hours (so TOT=40+OT), then w=OT/TOT*1.5*w_1+40/TOT*w_1. So, an employer who views a new worker as worth $10/hr (so w=10), and expects her to work 10 hours of OT per week, would offer her a base wage (w_1) of $9.09.”

[12] See footnote 5; this conclusion applies to the equilibrium outcome after employers have had time to adjust wages.

[13] Sachiko Kuroda and Isamu Yamamoto, “Impact of Overtime Regulations on Wages and Work Hours,” Journal of the Japanese and International Economies, Vol. 26, No. 2 (2014), pp. 249–262.

[14] Paul Davidson, “More American Workers Sue Employers for Overtime Pay,” USA Today, April 19, 2012, (accessed October 28, 2014).

[15] Global Workplace Analytics, “Latest Telecommuting Statistics,” (accessed October 28, 2014). These figures exclude the self-employed.

[16] News release, “The Wage and Hour Litigation Epidemic Continues,” Seyfarth Shaw, May 22, 2014, (accessed June 3, 2015).

[17] World at Work, “Survey of Workplace Flexibility 2013,” Figure 8, (accessed October 28, 2014). World at Work surveyed 566 companies about their human resources policies. Among employers who offered each workplace a flexibility option, World at Work found that 99 percent of exempt, but only 62 percent of nonexempt, employees could telework on an ad hoc basis; 95 percent of exempt, but only 48 percent of nonexempt, employees could telework at least once a month; and 98 percent of exempt, but 48 percent of nonexempt employees could telework at least one day a week, but not full time.

[18] Paul Davidson, “More American Workers Sue Employers for Overtime Pay,” USA Today, April 19, 2012, (accessed October 28, 2014).

[19] Workplace Flexibility 2010, “Flexible Work Arrangements: The Fact Sheet,” Georgetown University Law Center, (accessed October 28, 2014).

[20] Ibid.

[21] Daniel S. Hamermesh and Stephen J. Trejo, “The Demand for Hours of Labor: Direct Evidence from California,” Review of Economics and Statistics, Vol. 82, No. 1 (2000), pp. 38–47.

[22] Technically, Hamermesh and Trejo (2000) controlled for pre-existing trends with a difference-in-differences regression on changes in daily overtime rates between men in California and other states. They also used a difference-in-difference-in-differences regression including California women in the analysis as the third difference. California legislation already required daily overtime for female employees so the law did not affect their labor costs.

[23] Dora Costa, “Hours of work and the Fair Labor Standards Act: A Study of Retail and Wholesale Trade, 1938–1950,” Industrial and Labor Relations Review, Vol. 53, No. 4 (July 2000), pp. 648–664.

[24] Jane Friesen, “Overtime Pay Regulation and Weekly Hours of Work in Canada,” Labour Economics, Vol. 8, No. 6 (2002), pp. 691–720.

[25] David Bell and Robert A. Hart, “Wages, Hours, and Overtime Premia: Evidence from the British Labor Market,” Industrial and Labor Relations Review, Vol. 56, No. 3 (April 2003), pp. 470–480.

[26] John Johnson, “The Impact of Federal Overtime Legislation on Public Sector Labor Markets,” The Journal of Labor Economics, Vol. 21, No. 1 (2003), pp. 43–69.

[27] News release, “Union Members–2013,” Bureau of Labor Statistics, January 24, 2014, Table 3, (accessed October 31, 2014).

[28] Stephen J. Trejo, “Does the Statutory Overtime Premium Discourage Long Workweeks?” Industrial and Labor Relations Review, Vol. 56, No. 3 (April 2003), pp. 530–551.

[29] Mikal Skuterud, “Identifying the Potential of Work Sharing as a Job Creation Strategy,” Journal of Labor Economics, Vol. 25, No. 2 (April 2007), pp. 265–287.

[30] Sachiko Kuroda and Isamu Yamamoto, “How Are Hours Worked and Wages Affected by Labor Regulations? The White-Collar Exemption and ‘Name-Only Managers’ in Japan,” University of Tokyo Institute of Social Science Discussion Paper Series No. F-147, 2009.

[31] Sachiko Kuroda and Isamu Yamamoto, “Impact of Overtime Regulations on Wages and Work Hours,” Journal of the Japanese and International Economies, Vol. 26, No. 2 (2012), pp. 249–262.

[32] Stephen J. Trejo, “The Effects of Overtime Pay Regulation on Worker Compensation,” American Economic Review, Vol. 81, No. 4 (1991). pp. 719–740.

[33] Barkume, “The Structure of Labor Costs with Overtime Work in US Jobs.”


James Sherk
James Sherk

Research Fellow, Labor Economics