Who Benefits From the New Overtime Regulations? A Data Analysis ofthe U.S. Department of Labor's Assessment

Report Jobs and Labor

Who Benefits From the New Overtime Regulations? A Data Analysis ofthe U.S. Department of Labor's Assessment

August 16, 2004 35 min read Download Report
Visiting Fellow
Kirk Johnson
Former Visiting Fellow
Kirk is a Former Visiting Fellow.

The U.S. Department of Labor's new over­time regulation[1] will take effect on August 23. This regulation has broad implications for the more than 100 million workers who are sub­ject to the overtime provisions of the Fair Labor Standards Act (FLSA) of 1938. The fol­lowing is an economic analysis of who will be affected by the rules within the new regulation.

Specifically, this Center for Data Analysis (CDA) report concludes that:

  • Nearly 1.3 million low-income white-col­lar workers who do not currently receive overtime protections would become eligi­ble under the new regulation.[2]
  • Few workers who earn between $23,660 and $100,000 per year would be adversely affected by the new regulation. Those who currently enjoy overtime protections would keep them; and in many cases, those protections would be strengthened.
  • Some workers who earn $100,000 or more per year might lose their overtime protec­tions. While it is difficult to ascertain an exact number, between 108,000 and 163,000 salaried workers and between 98,000 and 147,000 hourly workers could be affected. Thus, at most, about 300,000 workers might lose their overtime protec­tions because of the new "highly compen­sated" test.
  • Therefore, a net 1 million workers will gain overtime protections through the new regulation.

Further, recent criticism from the Economic Policy Institute (EPI) that as many as 6 million workers would lose their overtime protections is considerably overstated.

  • Most of the workers that EPI argues would "lose their overtime" currently do not receive overtime because of the existing exemptions for executive, administrative, and professional workers.
  • The EPI analysis ignores or substantially discounts some new rules that favor the employees and make it more difficult for employers to deny overtime protections.
  • The EPI analysis is otherwise riddled with inaccuracies that have been previously identified (but ignored) by EPI.

Therefore, policymakers and the general public should discount EPI's recent analysis.


The FLSA generally requires most employers to pay workers a minimum wage and "time-and-a-half" premium pay when covered employees work more than 40 hours in any given work­week. The FLSA outlines in broad terms a num­ber of exemptions from these provisions, including one for "any employee employed in a bona fide executive, administrative, or profes­sional capacity…or in the capacity of outside salesman."[3] The FLSA also notes that these terms are "defined and delimited" by the Secretary of Labor from time to time.

To that end, on March 31, 2003, the Wage and Hour Division of the U.S. Department of Labor (DOL) published a notice of proposed rulemak­ing[4] to update the current regulation. During the three-month comment period, DOL received more than 75,000 comments. In addition, the proposal generated substantial debate in public policy cir­cles. After considering the various comments, DOL made a number of revisions and in the spring of 2004 published its new "final rule," which will take effect on August 23, 2004.[5]

Before discussing the economic effect of the new overtime regulation, a brief discussion of the rules is in order. An executive, administrative, or profes­sional worker must meet three tests before the worker can be considered exempt for overtime.

  • First, the "salary-basis" test specifies that a worker must be paid on a set salary. (Hourly employees who work more than 40 hours in a given workweek are generally eligible for overtime.)
  • Second, the "salary-level" test determines that any worker who earns less than a certain salary is automatically eligible for overtime, regard­less of the job.
  • Finally, a "duties" test determines which execu­tive, administrative, and professional employ­ees are exempt from overtime based on the kind of work they do.

The salary test has not been updated since 1975, and the duties test has gone even longer without any significant changes. One consequence of the outdated rules is that the minimum salary needed for exemption from overtime is an absurdly low $155 per workweek.[6] The new regulation would raise the salary test to $455 per workweek. Any worker paid less than $455 per workweek ($23,660 per year) would automatically receive overtime pro­tection, regardless of job duties.

There are separate duties tests for executive, administrative, and professional employees. While the new rules include a few changes in the lan­guage of the three duties tests, these changes gen­erally do little more than incorporate 50 years of court precedent-improving understanding and decreasing confusion among employers.

The current rules have a stringent "long duties" test for any executive, administrative, or profes­sional worker earning between $155 and $249 per week, and a "short duties" test for any worker earning at least $250 per week. Because the new salary-level test sets the minimum threshold at $455 per week, the long duties test no longer applies. Thus, the relevant changes are in the short duties test for each of the three classes of employ­ees. The current duties tests and the new duties tests are both shown in Table 1.

The new duties tests are more detailed than the current tests for a variety of reasons. First, the findings of 50 years of court cases have been inte­grated into the duties tests.[7] Second, some of the regulatory interpretation of the current rules has been integrated directly into the duties tests them­selves in order to minimize confusion.[8]

There are other separate duties tests specifically for computer employees, written to align with the standards set by Congress in 1996.[9] There is also a new duties test for highly compensated employees:

An employee with total annual compensation of at least $100,000 is deemed exempt under section 13(a)(1) of the Act if the employee customarily and regularly performs any one or more of the exempt duties or responsibilities of an executive, administrative or professional employee identified in subparts B, C or D of this part.[10]

"Total compensation" is then operationally defined as salary, commissions, nondiscretionary bonuses, etc. It does not include insurance premi­ums, retirement programs, or other fringe benefits. In short, a highly compensated employee can qualify as exempt under the overtime regulation by earning more than $100,000 per year and regu­larly performing the duties of an executive, admin­istrator, or other professional.


Data Analysis

This CDA report evaluates the most likely effects of the new regulation, when compared with the current regulation. Will workers who currently enjoy overtime protections lose them under the new rules? If so, how many and which kinds of workers will be affected? Conversely, will some workers benefit from the new regulation? In other words, are there workers who are currently exempt who will be reclassified as non-exempt because of the new rules?

To analyze both the proposed rule and the final rule, DOL hired a contractor, CONSAD Research Corporation of Pittsburgh. DOL's economic analy­sis of the final rule is published in the Federal Register.[11]

A number of observers have criticized CON­SAD's work on the proposed rule as difficult to replicate. For example, the AFL-CIO argued that "the study's methodology is confusing, and because CONSAD does a poor job of explanation, it is not capable of replication."[12] This CDA report seeks to conduct an economic analysis of the effects of the final rule, based on DOL's approach. To that end, this analysis uses the 2002 Current Population Survey's "Outgoing Rotation Group" database[13]-the same file that DOL, CONSAD, and other researchers have used for this purpose.

First, an estimate of the total employed U.S. population is required. There are about 134.3 mil­lion civilian workers aged 16 and older in Amer­ica. Not all of them receive overtime protection under the FLSA. The FLSA exempts certain kinds of workers from overtime protections, including the self-employed; federal workers, who are cov­ered under separate regulations administered by the Office of Personnel Management; certain trans­portation, agricultural, and mechanical workers; and clergy. (See Table 2.)


Although the values in this table are typically quite close to those reported in the Federal Register notice, there is a substantial difference in the size of the federal workforce. While Table 3-1[14] indi­cates that there are just over 1.5 million federal civilian workers, there are actually more than 2.35 million. Because of this 800,000-worker discrep­ancy, the overtime population reported in this CDA report differs from DOL's popula­tion. However, this error does not sig­nificantly affect the results because the FLSA does not cover federal workers.

As shown in Table 2, the CDA esti­mates that almost 114 million people are covered by the FLSA. Put another way, about 20.3 million are specifically exempted and are therefore not included in this analysis.

The DOL analysis subdivides this relevant population into five groups, based on the probability of exemption from the overtime rules: blue collar workers (not exempt from the overtime rules); those with a high probability of exemption (90- 100 percent); those probably exempt (50-90 per­cent); those probably not exempt (10-50 percent); and those with a low or no probability of exemp­tion (0-10 percent). These categories are based on the expertise of career staff from DOL's Wage and Hour Division and have been used by the U.S. General Accounting Office[15] and other research­ers.[16] (See Table 3.)


Next, the population is subdivided again into hourly and non-hourly workers. All hourly work­ers, regardless of job classification, are entitled to overtime protections because of the salary-basis test. Only non-hourly (salaried) workers are sub­ject to these regulations. (See Table 4.)


Three numbers in Tables 2, 3, and 4 warrant discussion. Of the 134.3 million civilian workers aged 16 and older, 20.3 million (15.1 percent) are not covered by the FLSA; 68.6 million (51.1 per­cent) are hourly workers; and the remaining 45.4 million (33.8 percent) are non-hourly workers.

These figures are similar to those in the Federal Register,[17] but they differ in two important respects.

First, the CDA analysis shows 800,000 addi­tional workers not covered by the FLSA because CONSAD/DOL undercounted the num­ber of federal workers.

Second, CONSAD/DOL double-counted about 800,000 private household workers. In their analysis, they included these observations twice, probably due to a file corruption problem. The vast major­ity of these twice-counted persons were either private household cleaners or pri­vate child-care workers (nearly 700,000). Since these two occupational categories are classified as blue-collar, the error does not substantially affect DOL's analysis.

However, because the above-men­tioned errors cancel out each other when calculating the total number of workers, CONSAD/DOL arrived at the target of 134.3 million total civilian workers. These discrepancies explain why most of the CDA results differ slightly from CON­SAD/DOL numbers.

The CONSAD/DOL analysis focuses on salaried white-collar workers when esti­mating who is exempt and non-exempt from the overtime rules. Eliminating blue-collar occupations from this analysis leaves about 71.9 million white-collar workers, who are paid either hourly or non-hourly. Of these, 33.8 million are hourly white-collar workers, and 38.2 million are non-hourly or salaried white-collar workers. This analysis therefore focuses on these 38.2 million salaried white-collar workers. Some of these white-collar occupations are specifically exempted from the salary test and are excluded from this analysis, as shown in Table 5.


Therefore, the relevant population is 106.4 mil­lion workers. Now that these intermediate steps have been taken, the total and exempted popula­tions can be estimated. The overtime rules are such that managerial requirements of jobs increase the likelihood of exemption from the overtime rules. There is a correlation, albeit imperfect, between salary and managerial duties (and, there­fore, exemption from overtime rules).

DOL has developed a probability model (based on the gamma distribution) to assign probabilities of exemption, based on the percentage categories discussed above. A gamma cumulative probability distribution function is one that best fits data that are skewed, as income data often are. Relatively few people earn great sums of money, while most workers earn lower amounts that tend to "bunch" together.[18]

When DOL and its contractor evaluated various probability model alternatives (linear, normal, log­normal, and gamma),[19] they determined that the gamma distribution best fits the data. However, their later results suggest that alternative models would have generated similar outcomes. Again, the aim is to evaluate the probability that a given salaried white-collar worker will be an exempted employee.

In applying the gamma probability model, individuals with low salaries are assigned the lowest probability of exemption within the specified cate­gories. Individuals with high salaries are assigned the highest probabilities. The 106.4 million workers are dis­tributed by method of payment (hourly/non-hourly), as shown in Table 6.


Thus, the population for this analysis consists of 31.7 million salaried white-collar workers. When the numbers are parsed in this manner, it becomes obvious that the vast majority of workers in America enjoy overtime pro­tections, either because they are paid by the hour or because they are in blue-collar occupations.

As noted, workers who earn less than $455 per week are auto­matically considered non-exempt, and workers who earn more than $100,000 per year (or $1,923.08 per week) are evaluated using the highly compensated test. Table 7 distributes these workers by the relevant earnings category.


Under the previous DOL rule, the more than 800,000 workers reporting weekly earnings of less than $155 per week are covered immediately under the old overtime rule. These individuals are therefore not in the probability analysis, but are included as a category in the analysis below.

When employing the gamma exemption proba­bility model, 18.7 million workers are estimated to be exempt under the overtime rules. They are dis­tributed, again, by salary classification, as shown in Table 8.


In summary, although the CONSAD/DOL analy­sis has some database problems, these do not sub­stantially alter the outcome of the analysis. Furthermore, because this analysis is interested in certain specified, salaried white-collar workers and the data problems are mostly confined to blue-collar occupations, these numbers are quite similar to those found in the Federal Register.[20]

Table 9 shows a summary of the results. In short, almost 1.3 million currently exempt, salaried white-collar workers would receive overtime protection under the new rule, by virtue of its newly promul­gated "bright line" salary test. This is consistent with what the CONSAD/DOL analysis found.[21]


The effect of the new "highly compensated" test is more elusive. As Table 9 shows, over 1.9 million of the 2.2 million highly compensated salaried workers are already exempt. That leaves only about 272,000 workers who might be made exempt under the new regulation. Some undoubtedly would stay non-exempt (and continue to collect overtime) if they work more than 40 hours in a given workweek. How many would be likely to do so is an open question.

Long-standing Supreme Court legal precedent has held that workers cannot be denied overtime just because they are highly compen­sated.[22] Indeed, the reason many of these workers command a high salary is that they are highly sought after and their employers therefore might have trouble retaining their services if they were sud­denly to deny them overtime. Perhaps 40 to 60 percent of these highly compensated workers (108,000 to 163,000 workers) might then lose their overtime protections.

The new highly compensated test might induce some employers to convert a por­tion of their hourly workers into salaried workers and classify them as exempt.[23] While ascertaining how many workers might be converted in this way is diffi­cult-if not impossible-determining how many hourly employees are paid more than $100,000 per year ($1,923.08 per week) ispossible. Of the 2.5 million highly compensated workers subject to the salary test, only 245,000 are paid by the hour, as shown in Table 10. Using the same standard to esti­mate how many of these workers might lose their overtime protection produces a range of 40 percent to 60 percent (97,000 to 146,000 workers).

However, these figures may be somewhat inflated. Analyzing the occupations of these highly compensated hourly workers reveals an interesting pattern. A majority of these workers (61 percent) is employed in an occupation with a high proba­bility of exemption from the FLSA. (See Table 11.) Most could arguably be converted under the cur­rent rule but have not been, probably because of market forces or other practical reasons.[24] There­fore, the estimated conversion figures are likely on the high side.

Although a total of 300,000 highly compen­sated workers might eventually lose their overtime protections, 1.3 million low-to-moderate earners would gain overtime protections. There­fore, a net of 1 million additional workers would enjoy overtime protections under the new regulation.

The EPI Study

The results of this CDA analysis starkly contrast to a report recently released by the Economic Policy Insti­tute. The EPI report suggests that the overtime regulation change "could strip away the right to overtime pay for over 6 million workers."[25] Given this sub­stantially different conclusion, the EPI report and the differences between this CDA report and the EPI analysis merit some examination.

As noted above, understanding how the new regulation differs from the old regulation is important in evaluating its effect. EPI argues that the changes in the regulation will cause certain workers to lose over­time protections. In most cases, the contention that millions of workers would lose their statutory overtime is unlikely because many of the changes on which EPI focused are minor, or even superficial.

Salary Test. As noted before, all salaried employees who are not specifically exempted from overtime by the FLSA (outside sales agents, teach­ers, mechanics, etc.)[26] and earn less than $455 per workweek would be eligible for overtime protec­tion. According to the CDA analysis, nearly 1.3 million new workers would be eligible for over­time protection, which is similar to the figure reached by the CONSAD/DOL analysis. EPI quib­bles with this number, arguing that DOL method­ology is unsound and includes in its calculations workers who are already exempt from the FLSA.[27]

Before evaluating the effect of the salary test, the CDA analysis specifically excluded those occupations exempted by statute, such as (among others) teachers, trans­portation employees, and outside sales workers. Then it estimated the number of exempt employees currently earning between $155 and $455 per week, using the probability model developed by DOL. The result is that nearly 1.3 mil­lion currently exempt salaried workers would enjoy overtime protection under the new regulation.

This does not mean that all 1.3 million workers would receive overtime pay each week-that depends on how many hours each employee works in a given week. The point of this analysis is to evaluate who is exempt and who is non-exempt from overtime protections, not to calculate the number of employees who work more than 40 hours per week. Elsewhere in their analysis, EPI evaluates overtime protections; but for some unknown reason, they deviate from that standard when evaluating the salary test.

Team Leaders. EPI argues that as many as 2.3 million "team leaders" will be exempted[28] under the new rule: "Section 541.203(c) exempts 'an employee who leads a team of other employees assigned to complete major projects for the employer.'" EPI argues that the term "major projects" is vague and could be interpreted to mean a variety of things.

However, in its regulatory interpretation section, DOL specifies that the kinds of high-level duties that would constitute such a "major project" include "purchasing, selling, or closing all or part of the business, negotiating a real estate transaction or col­lective bargaining agreement, or designing and implementing productivity improvements."[29]

EPI argues that employers could readily classify many employees into these kinds of projects. The problem with this analysis is that the current regu­lation is less restrictive than the proposed rule and allows employers to exempt "a wide variety of per­sons" who may do little else than "advise the man­agement."[30] Therefore, in this regard, the new rule strengthens, not weakens, overtime protections.

Pre-Kindergarten and Nursery School Teach­ers. EPI analysis suggests that 30,000 pre-kindergar­ten teachers would lose overtime protections under the new rule. Most teachers are already expressly exempt from the FLSA by statute, which DOL can­not change. Furthermore, the exemption of these teachers is not unique to the new regulation.

Those workers in child-care centers who also teach children, however, are not typically consid­ered exempted teachers. These workers' primary duty is ensuring the welfare of children in their charge, and teaching is a secondary duty. Therefore, their overtime protections are not in question.

"Rules of Thumb" for Working Supervisors. EPI contends that the "rules of thumb" that have long been used to help determine who is exempt from the overtime rules have been removed from the new regulation, making it easier "to reclassify supervisors as 'executives.'"[31] In fact, the rules of thumb are still in use. The new regulation simply calls them what they have always been-a "useful guide." Court cases have consistently held that there are situations in which managers can spend more than 50 percent of their time doing non-managerial work and still be exempt.[32]

A careful reading of the new duties test for exec­utives (see Table 1) clearly shows that the new duties test is more restrictive, thereby increasing the overtime protections of workers who might be misclassified as executives.

EPI also complains that line supervisors, who may do both production work and supervisory work simultaneously, may be more easily classified as exempt. Again, EPI is guilty of a selective read­ing of the new regulation. The Department of Labor's critique of a previous EPI study[33] states:

[T]he final rule retains the current regulatory requirement that an exempt employee's primary duty must be work directly related to the management or general business operations of the employer or the employer's customers, and includes a provision found only in the interpretive portion of the current rule (section 541.205(a)) clarifying that this phrase refers to activities relating to the running or servicing of a business as distinguished from working on a manufacturing production line or selling a product in a retail or service establishment.[34]

Therefore, in this case, the new regulatory rules governing working supervisors and line managers are more restrictive than the current ones. Again, the new "hire and fire" duties or "recommenda­tions … given particular weight" requirement added to the executive duties test make it more difficult to classify these employees as exempt executives.

EPI's final contention under this section is that some 548,000 hourly supervisors will be con­verted to a salary basis for the purpose of exempt­ing them from overtime protections. Given that the new rules are at least as stringent as-and, in some cases, more stringent than-the current rules, it seems unlikely that employers would rush to convert these employees.

Work Experience for Professionals.EPI argues that the new rules allow for work experi­ence or substitute knowledge to be used instead of a college or graduate degree and that this differ­ence would exempt some 900,000 workers.

Again, the current and new regulations are essentially identical; therefore, there is no change in the "course of study" standard for professional employees. Both state that a given occupation is "customarily acquired by a prolonged course of specialized intellectual instruction." This has been interpreted to mean that the individual has earned a college degree, but it is not limited to someone with that kind of credential. As DOL notes in the interpretive portion of the regulation, lawyers do not always go to law school,[35] chemists do not always have a college degree in chemistry, etc.

Both the current and new rules limit this narrow exception to those "who have substantially the same knowledge level and perform substantially the same work as the degreed employees."[36]

Additionally, EPI fails to properly document how they arrived at the 900,000-worker figure, further calling into question the legitimacy of this number. This figure could be based on the pro­posed rule change[37] that allowed for such "equiva­lent knowledge" to be attained in a number of ways. This language, however, was deleted from the final regulation.

Chefs and Sous Chefs. EPI argues that 130,000 chefs, sous chefs, and cooks will be exempted under the new rules. Once again, EPI fails to consider the effects of the current rules and new rules. The addition of a "creative" exemption only codifies and consolidates current case law on the subject. In fact, DOL takes great pains to delin­eate the currently exempt chefs who have culinary degrees from line cooks and other chefs whose duties are "predominantly routine, menial, man­ual, mechanical, or physical work."[38] The non-exempt status of these workers has been consis­tently upheld in court cases and DOL Wage and Hour Division opinion letters that go back more than 20 years.

Given the regulatory language in the rules, even if some sous chefs did lose their non-exempt sta­tus, the number would be nowhere near the 130,000 figure claimed by EPI.

Funeral Directors and Embalmers. Certain appellate decisions have allowed employers to deny overtime to funeral directors, embalmers, and other morticians who have as little as one year of training in mortuary science and two years of college.[39] The new rules specify that such an employee can be exempted only if he or she works in a state that requires four years of post-secondary education and a mortician's license, which cur­rently includes workers in about one-third of the states. Thus, overtime protections are strength­ened in this case, not weakened.

Financial Services Industry Employees.The majority of workers in financial services occupa­tions-especially tax experts, insurance agents, brokers, etc.-are already exempt because of the administrative exemption. What EPI fails to men­tion in its analysis is that this exemption is well established in case law, based on the current regu­lation.[40] Because the new regulation simply restates existing case law, the new rules are no dif­ferent from the old.

EPI is also confused about the nature of the financial services industry and how financial prod­ucts are marketed. Sales representatives in the financial services industry may be segmented into two types: active and passive sales agents. Active sales agents typically operate more in a consulting role-evaluating an individual's portfolio and financial needs; analyzing income, assets, family situation, and risk tolerance; suggesting alterna­tives; and finalizing transactions. These active sales representatives clearly qualify as exempt under both the new and current rules.

Others are passive sales representatives whose only job is to sell products, without the advice or analysis role. That is, they facilitate transactions rather than guide transactions through analysis of a client's personal information.

In more concrete terms, a stockbroker who works with a client to generate a financial plan may, in the course of that plan, suggest buying stocks or insurance contracts. That broker would be an exempt administrative employee. Another stockbroker who accepts unsolicited orders to buy or sell securities-and does not conduct financial analysis or consulting-would likely be non-exempt.

EPI argues that 160,000 mortgage loan officers will lose their rights to overtime because of the administrative duties test. Again, this is based on a selective reading of the new rules and disregards established case law that is not countermanded by the new regulation. The new rule adopts (from the current regulation) language established by case precedent.

For example, Casas v. Conseco Finance Corp.[41] held that mortgage loan originators were not exempt because they had a "primary duty to sell [the company's] lending products on a day-to-day basis" directly to consumers.[42] The court held that the loan originator performed mere production work and was consequently non-exempt. The new rules would continue these protections.

Computer Programmers. EPI laments the dele­tion of the "exercise of discretion and judgment" standard from the duties test. This standard was deleted in order to bring the regulation in line with the will of Congress, which rejected the discretion and judgment standard for computer professionals in FLSA Section 13(a)(17).[43] In short, the current DOL rule, as it pertained to computer program­mers, was extra-statutory. It should never have been put in the regulation in the first place and has therefore been deleted from the final version.

Even in the absence of the discretion and judg­ment standard, it would be difficult, if not impossi­ble, to find a computer programmer who does not exercise discretion and judgment in the course of writing computer code. As there invariably are mul­tiple ways to write a computer program, these pro­fessionals must by nature plan and structure code using their own judgment and experience. There­fore, the EPI argument is weak, and the operational effect of the new regulation is at most minimal.

Highly Compensated Employees. Highly com­pensated employees are the only workers who might lose overtime protection in significant numbers under the new rules. The question is how many workers would likely lose their overtime protection. EPI argues that the vast majority of all workers who earn more than $100,000 per year would lose over­time protection-some 400,000 in total.

Built into this argument is an assumption that a large group of hourly workers, who are non-exempt because they are not paid on a salary basis, would be converted into salaried employees and declared exempt. A good portion of these, accord­ing to EPI, would be registered nurses and phar­macists-occupations that would generally be classified as exempt, except that they are paid on an hourly basis. Market forces, not DOL regula­tions, determine the level and manner of pay for these highly sought-after workers.

To its credit, EPI does discuss the role of market forces in determining who might be converted from an hourly to a salaried basis. Because of these market forces, not all of these hourly workers would be converted to salaried workers. If between 40 percent and 60 percent of these hourly workers were eventually converted, that would mean that only 98,000 to 147,000 workers would lose overtime protection. However, that figure may be high, as noted in Table 11.

Of non-exempt salaried white-collar workers, approximately 272,000 are earning more than $1,923 per week ($100,000 per year). (See Table 9.) It is unlikely that all of these workers would be claimed as exempt, for two reasons.

First, even though EPI contends otherwise, the Supreme Court has long held that a worker cannot be claimed exempt just because he or she is well paid.[44]

Second, the duties test still stipulates that an exempt worker "customarily and regularly per­forms at least one of the exempt duties or respon­sibilities of an exempt executive, administrative or professional employee."[45]

EPI frets that the "customarily and regularly" standard is too vague and will lead to abuse, but the current rules use the same language. Further, the regulatory intent of the language clearly requires that these tasks be done much more frequently than the twice-per-year standard that EPI argues employ­ers will ultimately use to justify exempt status. (DOL elaborates and clarifies this point in its inter­pretive section of the new regulation.[46])

It is more reasonable to assume that 40 to 60 percent of these workers (108,000 to 163,000) would be reclassified. Thus, it is likely that rela­tively few workers would lose overtime protec­tions. Furthermore, the vast majority (nearly 90 percent) of salaried white-collar employees earn­ing $100,000 per year are already exempt under the current regulation.

Summary. EPI makes a host of unsubstantiated claims and unwarranted assumptions regarding the behavior of employers and employees, and this ren­ders their analysis suspect. In many cases, they are counting workers who supposedly "lose their over­time," even though the applicable section of the new regulation is operationally the same as the cur­rent regulation. In several cases, the rules are more stringent and favor the worker, making it more diffi­cult for the employer to deny overtime protection. On a number of levels, the EPI figures are suspect and should be viewed with a critical eye.

DOL also takes issue with the 2003 EPI report in a Federal Register notice,[47] pointing out a number of these same deficiencies. In a response to EPI's 2004 study, DOL concludes: "It is notable that EPI doesn't refute the Department's critique of their previous report on the proposed rule (69 FR 22210) and simply repeats the [sic] many of the same mistakes again in their latest and similarly flawed report."[48]

Finally, in contrast to the EPI analysis, this CDA report is grounded in the actual letter of DOL's new rules and analyzes how they compare to the old regulation.


The new overtime regulation is a welcome and long overdue update, and it has broad implications for the more than 100 million workers who are subject to the overtime provisions of the Fair Labor Standards Act. From an economic analysis of the data, this CDA report concludes that:

  • Nearly 1.3 million low-income white-collar workers who do not currently receive overtime protections would be eligible under the new regulation.
  • Few workers who earn between $23,660 and $100,000 per year would be adversely affected by the new regulation. Those who currently enjoy overtime protections would keep those protections-which would be strengthened in many cases.
  • Some workers who earn $100,000 or more per year might lose overtime protections. It is likely that 108,000 to 163,000 salaried work­ers and 98,000 to 147,000 hourly workers would be affected. Therefore, at most, 300,000 workers might lose overtime protections because of the new "highly compensated" test.
  • A net 1 million workers will gain overtime pro­tections through the new regulation.

Kirk A. Johnson, Ph.D., is Senior Policy Analyst in the Center for Data Analysis at The Heritage Foun­dation.


[1]U.S. Department of Labor, Wage and Hour Division, "Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales and Computer Employees," 29 C.F.R. Part 541, final rule, in Federal Register, Vol. 69, No. 79 (April 23, 2004), pp. 22212-22274.
[2]Estimates contained in this paper were developed from Current Population Survey data available from the U.S. Bureau of the Census. These data, along with the computer code used to generate the estimates, are available upon request.
[3]29 U.S.C. 213(a)(1).
[4]U.S. Department of Labor, Wage and Hour Division, "Defining and Delimiting the Exemptions for Executive, Administra­tive, Professional, Outside Sales and Computer Employees," 29 C.F.R. Part 541, proposed rule, in Federal Register, Vol. 68, No. 61 (March 31, 2003), pp. 15560-15597.
[5]Throughout this paper, there are references to the "current" and "new" regulations. The "new" regulation refers to the final rule, which will take effect on August 23, 2004. Until that time, the provisions of the "current" regulation govern employers and employees.
[6]To put this into perspective, a worker earning the $5.15 per hour minimum wage for 40 hours in a given workweek would earn $206, more than the current salary threshold of $155 per week.
[7]The myriad court cases were cited directly in the final rule and are not duplicated here.
[8]Because DOL's regulatory interpretation section was subject to the same public comment period and appears with the regulation itself, its interpretation has the same force of law. Integrating this interpretation directly into the duties test will increase transparency and decrease confusion among employers and employees.
[9]For the duties tests for computer employees, see U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22195.
[10]Ibid., p. 22269.
[12]Ibid., p. 22199.
[13]For a description of the outgoing rotation groups, see U.S. Bureau of the Census, Design and Methodology,Current Popula­tion Survey Technical Paper 63RV, pp. 10-12, at www.census.gov/prod/2002pubs/tp63rv.pdf (August 5, 2004). The outgoing rotation groups were used because they were the only individuals who were asked the earnings questions that are critical to this analysis.
[14]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22197.
[15]U.S. General Accounting Office, Fair Labor Standards Act: White-Collar Exemptions in the Modern Work Place, GAO/HEHS- 99-164, September 1999.
[16]For example, see Ross Eisenbrey and Jared Bernstein, "Eliminating the Right to Overtime Pay," Economic Policy Institute Briefing Paper, June 26, 2003.
[17]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22198, Chart 1.
[18]For a description of the gamma distribution in mathematical and visual terms, see National Institute of Standards and Technology, Engineering Statistics Handbook, Section, at www.itl.nist.gov/div898/handbook/eda/section3/ eda366b.htm (August 5, 2004).
[19]For a comparison of the various probability models, see U.S. Department of Labor, "Defining and Delimiting the Exemp­tions," final rule, p. 22205.
[20]Ibid., p. 22209, Table 3-7.
[21]Ibid., Chart 4.
[22]Jewell Ridge Coal Corp. v. United Mine Workers of America, Local No. 6167, 325 U.S. 161, 167 (1949).
[23]To meet the salary-basis test, an employer would first have to cease paying such employees by the hour in order to clas­sify them as exempt.
[24]Among these workers are some 61,000 registered nurses, pharmacists, and computer programmers who could be con­verted now under separate rules.
[25]Ross Eisenbrey, "Longer Hours, Less Pay: Labor Department's New Rules Could Strip Overtime Protection from Millions of Workers," Economic Policy Institute Briefing Paper, July 2004, p. 1.
[26]U.S. General Accounting Office, Fair Labor Standards Act, p. 41.
[27]Eisenbrey, "Longer Hours, Less Pay," p. 4.
[28]Ibid., pp. 4-5.
[29]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22264.
[30]Ibid., p. 22138.
[31]Eisenbrey, "Longer Hours, Less Pay," p. 6.
[32]For commentary and court citations on this subject, see U.S. Department of Labor, "Defining and Delimiting the Exemp­tions," final rule, pp. 22185-22186.
[33]Eisenbrey and Bernstein, "Eliminating the Right to Overtime Pay."
[34]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22211.
[35]For example, in Virginia, a person who did not go to law school may "read" for the bar exam via a program administered by the Virginia Board of Bar Examiners.
[36]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22265.
[37]U.S. Department of Labor, "Defining and Delimiting the Exemptions," preliminary rule, pp. 15560-15597.
[38]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22154.
[39]Ibid., pp. 22155-22156.
[40]See analysis in ibid., pp. 22145-22146.
[41]Casas v. Conseco Finance Corp., 2002 WL 507059 (D. Minn. 2002).
[42]For more commentary on the case, see U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, pp. 22145-22146.
[43]See commentary in ibid., pp. 22158-22160.
[44]See Jewell Ridge Coal Corp. v. United Mine Workers of America.
[45]U.S. Department of Labor, Wage and Hour Division, "Highly-Compensated Workers and the Part 541-Exemptions Under the Fair Labor Standards Act (FLSA)," Fact Sheet No. 17H, at www.dol.gov/esa/regs/compliance/whd/fairpay/ fs17h_highly_comp.htm (August 5, 2004).
[46]U.S. Department of Labor, "Defining and Delimiting the Exemptions," final rule, p. 22173.
[47]Ibid., pp. 22210-22212.
[48]U.S. Department of Labor, "The Economic Policy Institute Produces Another Flawed and Misleading Report," press release, July 14, 2004.


Visiting Fellow
Kirk Johnson

Former Visiting Fellow