Submitted electronically on November 7, 2023 via regulations.gov

Amy DeBisschop
Division of Regulations, Legislation, and Interpretation
Wage and Hour Division
U.S. Department of Labor
Room S-3502
200 Constitution Avenue NW
Washington, DC 20210

Re: RIN 1235-AA39, WHD-2023-0001, Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales, and Computer Employees

Dear Ms. DeBisschop:

The Economic Policy Institute (EPI) submits this comment in strong support of the Department of Labor’s Notice of Proposed Rulemaking Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales, and Computer Employees. The Economic Policy Institute (EPI) is a nonprofit, nonpartisan think tank created in 1986 to include the needs of low- and middle-income workers in economic policy discussions. EPI conducts research and analysis on the economic status of working America, proposes public policies that protect and improve the economic conditions of low- and middle-income workers, and assesses policies with respect to how well they further those goals.

The proposed rule would increase the salary threshold under which salaried executive, administrative, and professional (EAP) workers are entitled to overtime protections. This update will go a long way towards correctly restoring the value of one of the most precious resources workers have – their time.

The Fair Labor Standards Act of 1938 established bedrock labor standards for most workers in the United States. Enshrined in those protections is the right to a minimum wage, and the right to have a standard work week of 40 hours. Rather than capping or limiting the total amount of hours that could be worked, the Fair Labor Standards Act (FLSA) established overtime pay, meaning most workers have the right to be compensated at a rate of “time-and-a-half” – 1.5 times their typical rate of pay – for every hour worked over 40 in a week. A main purpose of overtime pay protections is to recognize that employers should bear some cost for overwork – they can make the choice to pay extra if they need their workers to work beyond 40 hours, or they can adjust workloads or staffing needs to fit a more reasonable 40-hour per week standard.

However, over time, the Department has failed to update the salary threshold frequently enough or sufficiently enough to ensure that the threshold is a meaningful proxy to correctly identify bona fide EAP-exempt workers. As described below, in 1975, 62.8 percent of full-time salaried workers were below the threshold and eligible for overtime, purely on the basis of the salary threshold. In 2023, even accounting for the 2019 rulemaking’s update to bring the threshold to $684 per week, merely 9 percent of full-time salaried workers are below that threshold and automatically eligible for overtime. As we will describe below, the Department’s proposed threshold is well within historical precedent in terms of the impacted share of the workforce. And the threshold could reasonably have been set even higher.

The Department’s 2019 updates to the regulations, which updated the salary threshold for exemption from $455 per year to $684 per year, have still left the threshold too low to meaningfully identify workers who should be exempt. $684 per week is the equivalent of just $35,568 per year for a full-year worker. This means many workers making just $36,000 per year – well below the income needed to provide a modest standard of living in many parts of the country – can be classified as overtime exempt as long as they have some EAP duties. And many other salaried workers will be misclassified as exempt – simply by virtue of being just over the threshold, their employers may be more likely to incorrectly determine that they pass the duties test, even if they don’t actually have the workplace leverage, discretion over their work, or the important and specialized training for job roles that bona fide EAP status is intended to identify.

In our analysis below, we provide several different points of comparison for the $1,059 per week salary threshold proposed by the Department, to explain why we believe that this proposed salary threshold is imminently reasonable, well within historic and economic precedent for the threshold for overtime exemption, and could reasonably be higher. We also describe the estimated impacts of the rule on the overall economy, including the impacted population of workers, and on employer costs and transfers to workers, which we estimate will be meaningful for those affected workers but represent a relatively small impact to the economy overall, meaning compliance with this rulemaking will be able to be borne easily by employers.

The proposed standard threshold could be much higher and still be well within historical precedent

The standard threshold could be substantially higher than what is proposed in the NPRM and still be well within historical precedent. The first way we show this is through a raw comparison of the proposed threshold to the inflation-adjusted values of earlier thresholds. Because of fundamental flaws in the threshold-setting methodology in 2004 and 2019, as explained below, we focus here on comparisons to earlier rulemakings.

One measure for comparison – adjusting past salary thresholds for inflation

The standard test is, as discussed below, essentially equivalent to the Department’s historic “short test” salary threshold, used prior to 2004, so it’s appropriate to compare the value of the standard threshold proposed here in 2023 to past values of the short test threshold. The 1975 short test threshold, $250 weekly (or $13,000 annually for a full-year worker), is equivalent to $1,304 (or $67,808 annually) in September 2023 dollars.1 Thus, the threshold proposed in the NPRM, $1,059, (or $55,069 annually) is 19% lower than the inflation-adjusted value of the 1975 short test threshold.

Looking further back: the threshold proposed in the NPRM is 25% lower than the inflation-adjusted value of the 1970 short test threshold ($1,418, or $73,760 annually); 20% lower than the inflation-adjusted value of the 1963 short test threshold ($1,321, or $68,698 annually); 9% lower than the inflation-adjusted value of the 1958 short test threshold ($1,166, or $60,619 annually); and 6% lower than the inflation-adjusted value of the 1949 short test threshold ($1,132, or $58,883 annually).2 The amount that the proposed threshold is below the average inflation-adjusted value of the short test for the five rulemakings between 1949 and 1975—the five rulemakings in which there was a short test—is 16%.

The standard test should arguably be somewhat lower than the historic short test to account for the fact that the elimination of the long test in 2004 means there is no longer a way for employers to exempt lower-paid salaried workers who would pass a rigorous duties test. However, 16% is a substantial decline that would more than account for the elimination of the long test. In other words, the threshold could be meaningfully higher than the proposed threshold and still be within historical precedent by this measure—and, as explained in the next section, this measure understates the effective difference between the threshold in the NPRM and earlier thresholds because it does not take into account changes in the wage and salary distribution over time.

A more rigorous empirical measure – the share of full-time salaried workers below the threshold

A more rigorous approach to showing that the threshold could be substantially higher than the proposed threshold and still be well within historical precedent is to look at the share of full-time salaried workers who earn less than the short/standard test threshold over time, and who are thus automatically eligible for overtime pay on the basis of salary alone. This is a more rigorous measure of how the proposed threshold stacks up against historic precedent than a comparison of the inflation-adjusted values of the threshold, because it accounts for changes in the wage and salary distribution over time.

In 1975, more than 60% (62.8%) of full-time, salaried workers earned less than the short-test threshold.3 Less than half of that share (28.2%) would earn less than the standard threshold proposed in the NPRM.4 The standard threshold could clearly be dramatically higher and still be well within historical precedent by this measure.

It is notable that the proposed threshold appears to be weaker relative to the 1975 threshold by this measure than by the simple comparison above of the proposed threshold level to the inflation-adjusted 1975 threshold level. This is due to the fact that wages and salaries grew substantially faster than prices in the intervening years,5 which means that in 1975, inflation-adjusted wages and salaries were substantially lower, resulting in a much larger share of the wage and salary distribution falling below the threshold in 1975. This measure provides a much better benchmark than a simple comparison of inflation-adjusted values, because it incorporates crucial information about changes in the wage and salary distribution over time.

It would be ideal if we could also look at the earlier years of rulemaking that involved the short test—1949, 1958, 1963, 1970—for further comparisons of this measure, but the necessary wage and salary data are not readily available. However, the above analysis of the inflation-adjusted values of the short-test for these years can provide insight. That analysis showed that adding in the earlier years did not dramatically change what was found by just looking at 1975. In particular, that analysis showed that the proposed threshold is 19% lower than the inflation-adjusted 1975 threshold, whereas it is 16% lower than the average of the inflation-adjusted thresholds from 1949, 1958, 1963, 1970, and 1975. Given that, we would not expect the share of full-time salaried workers earning below the short test threshold to be dramatically lower on average for the earlier years than it is for 1975. In fact, because, over the period of missing years, 1949-1970, wages and salaries grew substantially faster than inflation,6 we would expect the opposite. In the earlier years, inflation-adjusted wages and salaries were typically much lower, which would put strong upward pressure on the share of workers who earned less than the threshold in those years.

In other words, the 62.8% share of full-time salaried workers earning less than the short-test threshold in 1975 was not somehow a quirk of 1975, but instead the short-test years together (1949, 1958, 1963, 1970, and 1975) very likely averaged an even higher share. There is simply no doubt that the threshold could be at least as high as the proposed threshold (which only 28.2% of full-time salaried workers earn less than today) and still be well within historic precedent.

The proposed methodology is well within the bounds of historic methodologies

The preceding analyses showed that the proposed threshold could be much higher and still be within historic precedent by comparing how the proposed threshold stacks up against earlier thresholds given changes in the broader macroeconomic context. Now we turn to a comparison of how the methodology for calculating the proposed threshold stacks up against earlier methodologies. We find that here, too, the proposed threshold could have been much higher and still have been well within historic precedent.

Starting in 1949, the Department established two sets of tests for exemption, each of which involved a duties test and a salary test. One set of tests was the “long-test” set, which combined a stringent duties test (which included a 20% cap–40% in retail—on the amount of time that overtime-exempt employees could spend on nonexempt duties) with a lower salary threshold. The other was the “short-test” set, which combined a much more lenient duties test with a higher salary threshold. Thus, an employer who wanted to assert that a relatively low-paid employee was exempt had to show more rigorously that their duties were “bona fide executive, administrative, or professional” (EAP) in nature, whereas for a more highly paid employee, the employer did not have to make as rigorous a case.

In its 1958 rulemaking, the Department laid out a methodology, often referred to as the “Kantor” or “long test” methodology, for setting the long test threshold. In particular, the Department set the long test salary threshold so that only about 10 percent of employees performing EAP duties in the lowest-wage regions and industries would fail to meet the new salary level and therefore become entitled to overtime pay. This methodology was generally followed in its 1963 and 1970 rulemakings, while the 1975 rulemaking was intended to be temporary and was simply based on an adjustment of the 1970 threshold for inflation.

The short test salary threshold was always set significantly higher than the long test salary threshold to account for the fact that the short test threshold was meant to only impact those whose salary was high enough that it was very likely that they were bona fide EAP employees and as a result didn’t require as strict a duties test. From 1949 up until 2004, the short test salary level ranged from between 1.30 and 1.82 times the long test, with 1.49 being the average of the 15 historical ratios of the short test salary level to the long test salary level.

In the 2004 rule, the Department included just one set of tests. For this set of tests, the Department created a “standard” duties test that was essentially the more lenient “short test.” To remain consistent with the prior methodology, the Department should have paired this duties test with a higher salary threshold consistent with a short duties test. Instead, it used a lower salary level consistent with a long duties test. This was a fundamental error.

The fact that the 2004 threshold—$455 weekly or $23,660 for a full-year worker—was a pairing of the more lenient duties test with a long-test threshold is mentioned by the Department in numerous places, including in Table 6 of the 2004 final rule, where $455 is referred to as a long-test salary level; in Table 4 of the 2004 final rule, which shows that $455 is in fact somewhat lower than a long-test threshold, with only 8.2% of exempt workers failing to meet the $455 salary level in the South, less than the 10% previously used to set the long-test salary level; and in the 2017 request for information, in which the Department notes that “the $455 per week salary level was equivalent to the lower salary level that would have resulted from the methodology DOL previously used to set the lower long-test salary levels.”7 

Because of the mismatch between the duties test and the salary threshold in the 2004 rule, the methodology from the 2004 rule was not appropriate. Unfortunately, in the 2019 rulemaking, the Department repeated this error, .

As the Department reports in the NPRM, using current data, the Kantor, or “long-test,” methodology would result in a long test salary threshold of . Using the average of the 15 historic ratios of the short test to the long test, 1.49, the short test would be $1,378. As mentioned above, the standard test should be somewhat lower than the historic short test to account for the fact that the elimination of the long test in 2004 means there is no longer a way for employers to exempt lower-paid salaried workers who would pass a rigorous duties test. However, the Department’s choice of a standard threshold, $1,059,. The threshold could have been set substantially higher and still been well below the short test level.

The proposed threshold will affect—restore overtime protections to—a small share of workers

The number of workers affected by the current proposal is 3.6 million. It is a small share of the workforce – just 2.6% of workers subject to FLSA and the Department’s part 541 regulations, and just 2.3% of all workers.8^,9 However, it is important to note that the number or share of workers affected by any increase in the salary threshold is not an informative metric to use to assess whether any threshold is appropriate. The number or share of workers affected by a given threshold is the product of factors that are wholly unrelated to the appropriateness of the threshold, namely (1) whether the prior threshold was set at an appropriate level, and (2) how long it has been since the prior update in the threshold.

The longer it has been since the previous update, the more workers will be affected, as inflation and the overall wage and salary structure rise over time, eroding the effective level of the threshold. In the case of the current proposal, overall inflation has increased by 16.9% and the 35^th percentile of weekly earnings of full-time salaried workers in the lowest wage Census region has increased by since the threshold was last updated in 2019.10 Further, given that the 2019 threshold was set at an inappropriately low level, as described above, a large portion of the workers affected by the proposed, appropriately-set threshold would be affected simply to correct that fact. In other words, the workers who would receive new overtime protections if the proposed threshold were implemented would simply be having protections restored that they should have already been receiving.

In what follows, we use data available in the NPRM wherever possible. Where the NPRM did not provide data that we need for our analysis, we analyzed a dataset that we generated using the same source data used in the NPRM, 2020-2022 Current Population Survey Outgoing Rotation Group data, and by strictly following the methodology laid out in the NPRM.11

The number of workers affected by the proposed increase in the salary threshold (3.6 million) can be decomposed into two main components: workers affected as a result of correcting the error in the threshold-setting methodology from the 2019 update that resulted in a threshold that was too low, and workers affected by the erosion of the strength of the threshold since its last update as the overall wage and salary distribution has risen. There are 1.7 million workers affected as a result of correcting the error in the threshold-setting methodology from the previous update. This number is found by calculating the number of affected workers who earn less than what the overtime thresholds would have been had the 2019 rule set the threshold at an appropriate level, namely if it had been calculated by the methodology proposed in the current NPRM. In particular, it is found by summing the number of workers affected by the standard threshold who earn between $684 and $905 (where $905 is the 35^th percentile of weekly earnings of full-time salaried workers in the south in 2018), and the number of workers affected by the highly compensated employee (HCE) threshold who earn between $107,432 and $120,016, (where $120,016 is the 85^th percentile of weekly earnings of full-time salaried workers nationwide in 2018).12

There are 1.9 million workers affected by the erosion of the strength of the threshold since its last update as the overall wage and salary distribution has risen, which is similarly found by calculating the number of affected workers who earn at least what the overtime thresholds would have been in the 2019 rule if they had been calculated by the methodology proposed in the current NPRM—$905 for the standard test and $120,016 for the HCE test.

The number of workers affected as a result of correcting the threshold-setting methodology from the previous update, 1.7 million, is 1.2% of all workers subject to FLSA, and 1.1% of all workers. The number of workers affected as a result of the erosion of the strength of the threshold since its last update as the overall wage and salary distribution rises, 1.9 million, is 1.4% of all workers subject to FLSA and 1.2% of all workers. These are small, manageable shares that result in the restoration of overtime protections to workers who should have already been receiving them.

It is worth noting, of course, that while only a relatively small share of the overall workforce will be currently impacted, the rule as proposed would still be incredibly meaningful to the 3.6 million workers affected. In particular, we note that more than half of the workers impacted by the rule would be women. While women and Black and Hispanic men are less likely to be in impacted salaried EAP occupations overall, if they are in those occupations, due to systemic impacts of discrimination in the labor market, they are much more likely to be concentrated in lower-paying salaried positions, and to benefit from the expanded protections in this rule.

The proposal will not fundamentally alter the landscape of exemption

Another way to show the limited reach of the proposal is to look at the share of workers currently exempt from overtime compared to the share who would be exempt under the new proposal. It has seven data columns:

1. The number of workers subject to FLSA;
2. The share of workers subject to FLSA who are currently exempt from overtime under the 2019 threshold (calculated as the number of potentially affected workers divided by the share of workers subject to FLSA);
3. The share of workers subject to FLSA who would be exempt from overtime if the 2019 threshold had been set appropriately (calculated as the number of potentially affected workers minus the number of workers who are exempt as a result of the error in the threshold-setting methodology in the 2019 rule, and dividing that difference by the share of workers subject to FLSA)13;
4. The share of workers subject to FLSA who would be exempt from overtime if the 2019 threshold had been set appropriately and had not been allowed to erode since 2019 as the overall wage and salary distribution rose (calculated as the number of potentially affected workers minus both the number of workers who are exempt as a result of the error in the threshold-setting methodology in the 2019 rule and the number of workers who are exempt as a result of the erosion of the strength of the threshold since 2019, and dividing that difference by the share of workers subject to FLSA)14. This is also the share exempt under the current proposal.
5. The percentage point decline in exemption as a result of correcting the error in the threshold-setting methodology in the 2019 update (this is the difference between (2) and (3)).
6. The percentage point decline in exemption as a result of updating the threshold to account for the rise in the overall wage distribution since the 2019 update (this is the difference between (3) and (4)).
7. Total percentage point decline in exemption under the proposed threshold (this is the difference between (2) and (4)).

Overall, 20.3% of workers subject to FLSA are currently exempt from overtime protections. If the 2019 threshold had been set appropriately, that share would be 19.1%, and if the 2019 threshold had been set appropriately and had not been allowed to erode as the wage and salary distribution rose since that time, it would be 17.7%. Note, this 17.7% is also the share of workers subject to FLSA who would be exempt under the current proposal. So, 2.6% (= 20.3% – 17.7%) of workers subject to FLSA are currently exempt from overtime but would newly receive them under the proposal. This is a small, manageable share, and represents an important restoration of overtime protections to workers who should already be overtime eligible.

Table 1 does this breakdown for a wide range of worker categories: gender, race/ethnicity, age, education, industry, occupation, census division, and sector. One important finding is that under the current proposal, categories of workers in which a high share of workers are currently exempt from overtime would continue to have a high share of workers exempt from overtime under the proposal. In other words, the proposal does not exclude from exemption any categories of workers that have traditionally had a high rate of exemption.

The category of worker with the highest rate of exemption currently is workers in management, business, and financial occupations, at 60.9%. Under the new proposal, this category would see a decline of 6.8 percentage points, to 54.1%. This decline, the largest of any of category of worker, is nevertheless quite modest. The majority of it (3.7 percentage points) is simply to account for the overall rise in the wage and salary distribution since the threshold was last updated in 2019. The remainder (3.1 percentage points) accounts for the error in the threshold-setting methodology in the 2019 rule.

The category of worker with the next highest rate of exemption currently is workers in the financial activities industry, at 43.1%. Under the new proposal, this category would see a modest decline of 4.9 percentage points, to 38.3%. Again, the majority of this decline (2.6 percentage points) is simply to account for the overall rise in the wage and salary distribution since the threshold was last updated in 2019. The remainder (2.2 percentage points) accounts for the error in the threshold-setting methodology in the 2019 rule.

The category of worker with the third highest rate of exemption currently is workers in professional and business services, at 40.2%. Under the new proposal, this category would see a modest decline of 4.1 percentage points, to 36.1%. Again, the majority of this decline (2.3 percentage points) is simply to account for the overall rise in the wage and salary distribution since the threshold was last updated in 2019. The remainder (1.8 percentage points) accounts for the error in the threshold-setting methodology in the 2019 rule.

In the remaining categories, current exemption ranges from 0.1% to 39.6%, and the decline in exemption under the new proposal ranges from zero to 4.7 percentage points, with majority of the decline in most categories being simply to account for the overall rise in the wage and salary distribution since the threshold was last updated in 2019. Table 1 underscores that the proposed threshold won’t fundamentally alter the landscape of exemption in any major worker category, and is also a reminder that the modest declines in exemption that would occur if the proposal were finalized would be the result of restoring overtime protections to those workers who should already be receiving them.

The proposed threshold will have a minimal impact on the broader economy

We have shown in many ways that while the proposed overtime protections would matter a great deal to those who receive these new protections, this proposal, if finalized, would not have a large impact on the broader economy. Another way to demonstrate this is to consider the transfers affected workers will receive as a part of this rule relative to wage and salary benchmarks from the broader economy.

The total “wage bill” in 2022—that is, the total amount of wages and salaries paid to workers in the U.S. economy—was $10.5 trillion.15 The total transfers in the first year of implementation if the proposal is implemented is estimated to be $1.2 billion. These total transfers are 0.012% of the total wage bill—far less than one-tenth of one percent (in fact, just slightly above one-hundredth of one percent). This is a tiny share of the overall wage bill, underscoring that while this proposed rule would be meaningful to those who receive new protections, it will have a barely noticeable impact on the broader economy.

Even when direct employer costs for the first year of implementation—an additional $1.2 billion—are added in, the total transfers and direct employer costs are just 0.023% of the total wage bill—again, far less than one-tenth of one percent (slightly above two-hundredths of one-percent). And of course, direct employer costs drop dramatically after the first year of implementation, as costs are just an estimate of the expenses associated with coming into compliance, rather than increased wages for workers. Again we see that the threshold could have been set substantially higher and still not cause a major disruption to the economy or to employers.

The proposed threshold maintains the importance of the duties test

Another important point regarding the appropriateness of the proposed threshold is that it maintains the importance of the duties test. To be exempt from overtime, a worker must be paid on a salary basis, earn above the salary threshold, and pass the duties test. If the threshold is set so high that very few workers who earn above it actually fail the duties test, then the threshold will have effectively rendered the duties test moot. This is not at all the case with the proposed threshold. In fact, as shown in Figure B in the NPRM, more than half of white-collar, salaried workers who fail the duties test earn above the proposed threshold (in particular, 7.6 million white-collar, salaried workers earn above the proposed threshold, while 7.5 million earn below). In other words, more than half of white-collar salaried workers who should be eligible for overtime under the proposal have their overtime eligibility determined entirely by the duties test.

Further, a sizeable portion—nearly one-in-four (7.6/32.1 = 24%)—salaried white-collar employees who earn above the threshold fail the duties test, underscoring that under the proposed threshold, the duties test continues to play a highly important role in the determination of exemption. It is worth noting that if the threshold were set at the inappropriately-low long-test level, $925, that share rises only to 26.1%.16 In other words, the degree of relevance of the duties test under the proposed threshold is clearly not too low—it remains very much in line with what it would be even if the threshold were set at the inappropriately-low long-test level.

Further, it would be the case that the importance of the duties test would be maintained even at substantially higher thresholds. For example, if the threshold were set at the short-test level, $1,378, a still-sizeable portion—nearly one-in-five (19%)—salaried white-collar employees who earn above that threshold would fail the duties test. The threshold could have been set much higher and the duties test would have retained a crucial role.

If the salary threshold remains too low, the Department risks the continued status quo of placing too much importance on the duties test to determine exemption. Application of the duties test by individual employers is highly variable, as whether or not a given employee passes the duties test must be determined by the reality of their actual job responsibilities, rather than on job descriptions or on clearer or more standardized characteristics such as job titles. Every application of the duties test puts the worker at risk of misclassification as EAP-exempt, due to different employers’ partially subjective applications of the duties test. For the salary threshold to be effective, it must be high enough to ensure that workers with lower salaries are automatically entitled to overtime protections, so that only workers earning over a reasonably high amount will be subject to the duties test to determine their overtime eligibility.

Automatic updating maintains the proposed standard much more effectively than a static threshold

Automatic updating of the threshold is crucial. If the threshold is not updated automatically over time, it will steadily weaken as a labor standard until the next rulemaking, covering fewer and fewer workers as the salary distribution naturally rises over time with inflation and productivity growth. A static threshold would get further and further away from the standard laid out in the proposal over time, whereas a threshold updated in the manner proposed in the NPRM would preserve the standard laid out in the proposal until the next rulemaking.

The proposed automatic updating is done in a way that exactly maintains the standard the proposal establishes. The updates would simply reset the threshold to the exact standard laid out in the proposal, the 35th percentile of weekly earnings of full-time salaried workers in the lowest-wage Census region. As the 35th percentile of weekly earnings of full-time salaried workers in the lowest-wage Census region rises over time, the threshold would rise with it, ensuring that the standard laid out in the new rule is simply preserved – neither strengthened nor weakened.

Automatic updating is a smart and easy way to simply maintain the labor standard established in the proposal. Automatic updating preserves the status quo while leaving open the opportunity for policymakers to change the substance of the rule in a future rulemaking, if and when they believe such a change to be necessary.

It is not reasonable to assert that the Department should just do more frequent rulemaking instead of automatic updating. Rulemaking is extremely time and resource-intensive, taking thousands of hours of work by dozens of staff. It doesn’t make sense to continually devote limited resources just to maintain the status quo in a regulation when it can be done automatically. Notice and comment rulemaking should be reserved for when policymakers want to change the substance of the rule.

What’s more, automatic updating provides crucial predictability for employers. With automatic updating, employers will know exactly what to expect and when to expect it. They will also be able to get a reasonable sense of what the next threshold will be even relatively far out, because they will be able to track how the 35^th percentile of full-time salaried worker earnings in the lowest-wage Census region is evolving over time at a dedicated Bureau of Labor Statistics website.

An example of how this automatic indexing might have played out in recent years is informative. Assume the threshold had been set in 2019 to the 35^th percentile of weekly earnings of full-time salaried workers in the lowest wage Census region using 2018 data, and then updated annually. The threshold would have been set to $905 in 2019, $960 in 2020, $1,000 in 2021, $1,000 in 2022, and $1,059 in 2023. These are reasonable, absorbable increases—and notably, these reasonable increases are what would have occurred during one of the most volatile periods in modern economic history, the COVID-19- induced recession and its aftermath, during which inflation and wage growth reached levels not seen in well over a generation. That the proposed automatic updating would have generated such a reasonable growth pattern during the COVID recession and its aftermath means the proposed automatic updating passes one of the most extreme “stress tests” that one could conceive of.

It’s worth noting that in general, salaries grow faster when the economy is strong, and grow more slowly when the economy is weak.17 This means the automatic updating proposed in the NPRM would naturally create a business-friendly growth pattern. Namely, the automatic updates would tend to be larger when the economy is strong and businesses are, as a result, better able to absorb larger increases, and the automatic updates would be smaller when the economy is weak and businesses may be more likely to be feeling pinched. Automatic updating to maintain the threshold at the 35^th percentile of the weekly earnings of full-time salaried workers in the lowest wage Census region is good for employers and it protects workers. It is the very definition of efficient, effective government.

The threshold should be updated annually, not triennially

The Department proposed updating the threshold every three years. It would be much more effective to increase the threshold annually. An annual update would mean that the threshold more closely adheres to the 35^th percentile of the weekly earnings of full-time salaried workers in the lowest wage Census region, rather than slipping further and further behind in between triennial updates and then having an unusually large “catch up” increase.

If the threshold is increased annually, as we believe it should, then calculating the threshold using the most recent 12 months of data when an update is announced is reasonable. However, if the Department retains their proposal to increase the threshold every three years, then the threshold should be calculated using the most recent three months of data. The most recent three months of data will be, on average, higher than the most recent twelve months of data, and therefore is preferred under triennial updating, given that under triennial updating, the salary threshold would be suppressed (relative to annual updating) for two out of every three years. Further, while there could potentially be concern that quarterly data would be too volatile under annual updating, that would not be the case under triennial updating, where the three-year gap would smooth any excessive volatility.

The Department has clear authority to automatically update the threshold

Furthermore, the Department has clear statutory authority to update the salary threshold and to establish automatic adjustments to the threshold. On previous rulemakings around the EAP exemption, industry groups such as the U.S. Chamber of Commerce have pointed out that they believe automatic updates to the salary threshold are outside of congressional and statutory intent with regards to the DOL’s authority. However, updating the salary threshold, and establishing an automatic update to the threshold every 3 years, fully complies with the Administrative Procedure Act and the FLSA. Just as the Department has authority under 29 U.S.C. 213(a)(1) to establish the salary level test, DOL likewise has authority to automatically update the salary level to ensure it remains effective.

We support the rule’s proposal to establish the same standard salary threshold for the U.S. territories

EPI also supports the Department’s proposal to increase the salary thresholds that apply to the U.S. territories, and to apply the same threshold to all territories that are also subject to the federal minimum wage. This proposal is in keeping with typical practice in the Department, which has applied just the standard salary threshold to Puerto Rico, Guam, and the U.S. Virgin Islands. This proposal also acknowledges that the 2019 overtime rule’s decision to maintain a lower threshold even in territories subject to the federal minimum wage – Puerto Rico, Guam, the U.S. Virgin Islands, and the Commonwealth of the Northern Mariana Islands (CNMI) – was a deviation from previous precedent since the 2004 rulemaking. The proposed rule also acknowledges that the standard salary level will eventually also apply to American Samoa, pursuant to the Fair Minimum Wage Act of 2007.

We believe a standard salary threshold applying to all U.S. territories is not only within the precedent of previous rulemakings, but is also necessary for basic fairness and application of our fundamental labor standards. Workers in the U.S. territories should not be locked into artificially lower wage floors or to lower labor standards, and if the salary thresholds for exemption are not also raised substantially from the current levels set in 2004, those workers will be more at risk of having the value of the protection erode over time.

Increasing the salary threshold to the standard level in the territories is also an important step to ensure racial equity in the application of our labor standards. According to the 2020 Census, 90.3% of the population of the 18 According to the 2020 Island Areas Censuses, which includes surveys of households in American Samoa, the CNMI, Guam, and the U.S. Virgin Islands, in 2019, all 4 of those territories reported poverty levels significantly higher than the national levels for the 50 states and District of Columbia.19 While there are other factors contributing to these disparities in household income and pay, and many more to the longstanding economic hardship and repression present in the U.S.’s territories, establishing the same salary threshold for the territories would be an important step in ensuring that our labor laws do not continue to enshrine a segregated, multi-tier wage system that unfairly impacts Asian, Black, Hispanic, and Pacific Islander workers.

Expanding overtime protections is manageable for employers and good for workers

Employers will have plenty of flexibility for how they choose to respond to this rule – and more incentive to do so. Given that the current cost to employers of overworking lower-paid, salaried workers and managing them inefficiently is effectively zero, employers who find themselves with newly overtime-eligible employees will have more incentive to better manage their time. They may more carefully monitor workloads, pay time-and-a-half when they need their workers to go beyond 40 hours in a week, or even hire additional employees to cover heavy workloads and reduce the need for overtime pay. Some employers may also determine that they really do require a given salaried employee with bona fide EAP duties to continue working more than 40 hours a week regularly enough that they would rather just give them a raise to over the new threshold – meaning those employees may still work longer hours, but they will still get an increase in pay.

Overall, expanding overtime protections will mean more pay or more time back for employees who would otherwise be overworked. While this proposed rule would affect a relatively small share of the overall US workforce, it would make a significant impact to the lives of the workers who would gain new protections under the rule. We urge the Department to finalize and promulgate this rule without further delay.

Sincerely,

Heidi Shierholz, Ph.D.
Executive Director
Economic Policy Institute

Samantha Sanders
Director of Government Affairs & Advocacy
Economic Policy Institute

Jori Kandra
Research Assistant
Economic Policy Institute

---

1. Bureau of Labor Statistics, All Urban Consumers, Consumer Price Index Research Series (BLS-CPI-U-RS). Various years, public data series accessed through the CPI database. Wages are adjusted into September 2023 dollars by the CPI-U-RS using the annual value for 1975 and seasonally adjusted monthly value for September 2023.

2. Bureau of Labor Statistics, All Urban Consumers, Consumer Price Index Research Series (BLS-CPI-U-RS). Various years, public data series accessed through the CPI database. Wages are adjusted into September 2023 dollars by the CPI-U-RS using the annual value for 1949, 1958, 1963, and 1970, and seasonally adjusted monthly value for September 2023.

3. McNicholas, Celine, Samantha Sanders, and Heidi Shierholz. 2017. What’s at stake in the states if the 2016 federal raise to the overtime pay threshold is not preserved—and what states can do about it. Economic Policy Institute, November 2017.

4. Economic Policy Institute (EPI) analysis of Current Population Survey Outgoing Rotation Group (ORG) microdata, EPI Current Population Survey Extracts, Version 1.0.44 (2023), https://microdata.epi.org.

5. Bureau of Labor Statistics, Nonfarm Business Sector: Real Hourly Compensation for All Workers [COMPRNFB], retrieved from FRED (BLS-FRED). Public data series accessed through the Federal Reserve Bank of St. Louis.

6. Bureau of Labor Statistics, Nonfarm Business Sector: Real Hourly Compensation for All Workers [COMPRNFB], retrieved from FRED (BLS-FRED). Public data series accessed through the Federal Reserve Bank of St. Louis.

7. Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales and Computer Employees [final rule], 69 Fed. Reg. 22121 (April 23, 2004); Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales and Computer Employees [request for information], 82 Fed. Reg. 34616 (July 26, 2017).

8. Throughout this comment, we shorthand “subject to FLSA and the Department’s part 541 regulations” as “subject to FLSA.”

9. Defining and Delimiting the Exemptions for Executive, Administrative, Professional, Outside Sales, and Computer Employees, 69 Fed. Reg. 62152-62240 (September 8, 2023). According to Figure 1 of the NPRM, in 2022 the number of workers subject to FLSA was 139.4 million.

Bureau of Labor Statistics, Current Population Survey (BLS-CPS). 2022. Public data series accessed through CPS database. According to the BLS-CPS, in 2022 the number of wage and salary workers was 158.3 million.

10. Inflation is calculated from the CPI-U-RS, and the change in the 35^th percentile of weekly earnings of full-time salaried workers in the lowest wage Census region is calculated from an EPI analysis of Current Population Survey Outgoing Rotation Group microdata. In both cases, the base year was 2018 and the “current” year was 2022. This was done to account for the fact that, for example, a rule implemented in a given year would likely set the threshold using data from the prior year, or some portion of the prior year.

11. Though we followed that methodology as precisely as possible, we weren’t able to exactly match the Department’s numbers. For example, the Department found that 3.64 million workers would be affected by the proposal, whereas we found 3.86 million. To more closely benchmark the Department’s numbers, we reweighted our data so that our breakdowns by industry match the Department’s numbers in Table 9 in the NPRM.

12. We used 2018 data in the calculation of what the 2019 thresholds would have been had they been set appropriately, since a rule implemented in a given year would likely set the threshold using data from the prior year, or some portion of the prior year.

13. The calculation of the number of workers who are exempt as a result of the error in the threshold-setting methodology in the 2019 rule is described in the prior section. For “all,” the number is 1.7 million.

14. The calculation of the number of workers who are exempt as a result of the erosion of the strength of the threshold since 2019 is described in the prior section. For “all,” the number is 1.9 million.

15. U.S. Bureau of Economic Analysis, Total wages and salaries, BLS [BA06RC1A027NBEA], retrieved from FRED (BLS-FRED), Federal Reserve Bank of St. Louis.

16. Calculated from Figures B and C in the NPRM. Together, these figures show that 1.6 + 1.6 + 7.6 + 24.5 = 35.3 million white collar salaried workers earn above $925, and of those, 1.6+7.6=9.2 million, or 26.1%, fail the duties test.

17. See Bivens, Josh, and Ben Zipperer. 2018. The Importance of Locking in Full Employment for the Long Haul. Economic Policy Institute, August 2018.

18. U.S Census Bureau. 2020. ”General demographic characteristics.” Decennial Census of Island Areas, Commonwealth of the North Marina Islands.; U.S Census Bureau. 2020. ”General demographic characteristics.” Decennial Census of Island Areas, Guam; U.S. Census Bureau. 2020. Puerto Rico: 2020 Census (web page). Last modified July 12, 2023.; U.S Census Bureau. 2020. ”General demographic characteristics.” Decennial Census of Island Areas, American Samoa.; U.S Census Bureau. 2020. ”General demographic characteristics.” Decennial Census of Island Areas, U.S. Virgin Islands.

19. U.S. Census Bureau. 2023. ”2020 Island Areas Census Press Kit“ (news release). July 20, 2023.