Director, Product Classification
U.S. Postal Service
475 L’Enfant Plaza S.W., Room 4446
Washington, DC 20260-5015
PCFederalRegister@usps.gov 

Re: Proposed Rule: Ballot Mail for Federal Elections

The Economic Policy Institute (EPI) is a nonprofit, nonpartisan think tank that for 40 years has centered working families in economic policy discussions. EPI is submitting these comments in response to the Postal Service’s proposed rule on Ballot Mail for Federal Elections,1 which would have a disparate impact on Americans who face barriers to voting in person, including workers with disabilities, working parents, and workers with long and unpredictable work shifts. For this and other reasons outlined below, we believe that the proposed rule should be abandoned permanently and in its entirety.

The proposed rule follows a March 31, 2026, executive order from President Trump2 that would require the Postal Service to set new standards for the design of ballot envelopes used by state and local jurisdictions to facilitate centralized tracking of ballots to and from individual voters, thus encroaching on the authority to regulate and administer elections that the Constitution assigns to Congress and the states. The proposed rule also directs the Postal Service to compile a national voter list from state voter rolls and to reject ballots addressed to voters who are not on the list or that do not conform to the new envelope standard.

Tellingly, the proposed rule does not instruct the Postal Service to notify voters whose ballots were not delivered so that voters can challenge these decisions and correct errors caused by typos and similar discrepancies, which are vastly more common than deliberate fraud. Discrepancies and gaps in government records are not purely random, but are more likely to affect people with uncommon or hyphenated names (including many foreign-born citizens), married women who changed their names, and elderly and low-income Americans, among others.3

The proposed rule would misuse government data for political purposes.

The proposed rule should be viewed in the larger context of actions taken by this administration to use government data for unauthorized purposes, including voter suppression.

In addition to directing the Postal Service to compile a list of registered voters and use it to restrict mail voting, other provisions of the president’s executive order direct the United States Citizenship and Immigration Services (USCIS) and the Social Security Administration (SSA) to compile lists of voting-age citizens in each state, even though there is no evidence that fraudulent voting by noncitizens is a problem in U.S. elections.4 However, purging voter lists of suspected noncitizens could be used to disenfranchise eligible voters.

A case pending before the U.S. Supreme Court that would weaken the National Voter Registration Act could enable voter purges of suspected noncitizens close to elections when voters have little time to challenge errors that are common in such purges.5 Since some groups are more likely to vote for particular parties, purges can be weaponized for partisan advantage, a problem that would be magnified if done on a national scale.

The executive order adds to previous efforts by this administration to use SSA and other sensitive personal data for purposes beyond their intended use. It also risks another data breach in violation of federal privacy laws similar to an earlier breach of SSA data by a DOGE operative.6

The proposed rule would discourage voting by mail and harm working families who are more likely to face barriers to in-person voting.

Importantly, the harm inflicted by the proposed rule would extend beyond eligible voters who are directly prevented from voting by mail because they do not appear on the Postal Service list of registered voters. By casting doubt on the integrity and impartiality of mail voting, the rule would increase the number of voters dissuaded from voting by mail who later find themselves unable to vote in person.

Thirteen states, along with Puerto Rico and the Virgin Islands, restrict absentee voting to voters who know they will be out of the county on election day, or, in some states, who face barriers related to age, health, disability, work schedules, or other conflicts, such as jury duty.7 In my personal capacity as a volunteer on a voter assistance hotline, I can attest that many, if not most, people who face barriers to in-person voting could not have predicted them in advance. Voters frequently fall sick, face long lines at the polls that threaten to make them late for work, or find themselves with last-minute childcare and transportation problems.

Even voters who face predictable barriers that are valid reasons for absentee voting in their state can find it difficult to determine whether they qualify since specifics are not spelled out or are buried in dense legal language. What counts as a disability? Is documentation required? What if an anticipated work shift, jury duty, or vacation does not happen?

As the Institute for Policy Studies has pointed out, working-class voters are more likely to face barriers to voting in person due to work and family obligations.8 The Shift Project at the Harvard Kennedy School has documented the large number of low-wage workers, disproportionately workers of color, who work long and unpredictable shifts with little input into their schedules.9 In-person voting hours vary by state, but typically span a 12- or 13-hour time period.10 For working parents transporting children to school, workers with long commutes, and workers who face mobility challenges, it can be difficult if not impossible to vote in person within the designated window, especially if lines at the polls are long. Though some white-collar workers face these challenges, low-wage workers are less likely to work from home, have predictable schedules, or be given flexibility by employers to vote.

The Postal Service should scrap the proposed rule.

The above-mentioned constitutional, voting rights, and logistical problems with the proposed rule have been described in lawsuits and in commentary from a wide range of stakeholders and perspectives, including Lawfare,11 the Cato Institute,12 and the American Postal Workers Union.13 Twenty-three states and the District of Columbia successfully sued to temporarily block the executive order on which the proposed rule is based.14 However, the administration has said they will challenge the ruling, and regardless the rule could still take effect after the upcoming November election (the focus of the temporary injunction).

EPI believes that the proposed rule should be abandoned permanently and in its entirety for the following reasons:

• It is an unlawful attempt by the executive branch to seize control of elections from states and Congress.15
• It would inflict reputational damage on the Postal Service by involving it in decisions about who can and cannot receive ballots and vote by mail.16
• It would impose financial and logistical burdens on the Postal Service, which is already stretched to its limit.17
• It could jeopardize the timely delivery of all mail ballots, including those that conform to the requirements of the rule.18
• In combination with other provisions of the executive order, it could facilitate systematic voter purges for partisan advantage.
• It would cast doubt on the integrity and impartiality of mail voting.
• It would dissuade eligible voters from voting by mail, many of whom will face barriers to voting in person.

The Postal Service is an independent agency that, by design, is not under the direct control of the president and therefore not subject to his executive order.19 It has impartially delivered mail ballots to voters since the Civil War, when states introduced absentee voting for soldiers, a right later extended to other absentee voters. Some states have also extended vote by mail to voters who face specific barriers to in-person voting, such as people with disabilities. Other states and the District of Columbia have gone much further, mailing ballots to all registered voters. This is by far the fairest solution, but until it is the law of the land, we should work to extend, not restrict or suppress, mail voting.

Respectfully submitted,

Monique Morrissey
Senior Economist

Endnotes 

1. “Ballot Mail for Federal Elections: A Proposed Rule by the Postal Service on 06/02/2026,” Federal Register Published Document: 2026-10968 (91FR 32915). https://www.federalregister.gov/documents/2026/06/02/2026-10968/ballot-mail-for-federal-elections

2. Donald J. Trump, “Ensuring citizenship verification and integrity in federal elections,” March 31, 2026. https://www.whitehouse.gov/presidential-actions/2026/03/ensuring-citizenship-verification-and-integrity-in-federal-elections/

3. Monique Morrissey and Daniel Costa, “Cleaning up administrative records or targeting immigrants?” https://www.epi.org/blog/cleaning-up-administrative-records-or-targeting-immigrants/ 

4. See, for example, Wren Orey, Theresa Cardinal Brown, Feyisayo Oyolola, and Theo Menon, “Four Things to Know about Noncitizen Voting,” Bipartisan Policy Center, February 20, 2026. https://bipartisanpolicy.org/article/four-things-to-know-about-noncitizen-voting; Michael Waldman, “Why the Myth of Noncitizen Voting Persists,” Brennan Center for Justice, August 21, 2024. https://www.brennancenter.org/our-work/analysis-opinion/why-myth-noncitizen-voting-persists. Stephen Richer, “Trump’s Claims About Noncitizens Voting Are False. We Can Prove It.” Cato Institute, February 5, 2026. https://www.cato.org/commentary/trumps-claims-about-noncitizens-voting-are-false-we-can-prove-it

5. Jim Saksa, “Supreme Court will hear Arizona case that could badly weaken key federal law protecting voter registration,” Democracy Docket, June 29, 2026. https://www.democracydocket.com/news-alerts/supreme-court-will-hear-arizona-case-that-could-badly-weaken-key-federal-law-protecting-voter-registration/

6. Stephen Fowler and Jude Joffe-Block, “The Trump administration admits even more ways DOGE accessed sensitive personal data,” Weekend Edition, National Public Radio, January 30, 2026. https://www.npr.org/2026/01/23/nx-s1-5684185/doge-data-social-security-privacy

7. National Council of State Legislatures, Table 2: Excuses to Vote Absentee, website accessed July 2, 2026. https://www.ncsl.org/elections-and-campaigns/table-2-excuses-to-vote-absentee

8. Sarah Anderson, “Attacks on Mail Voting are Attacks on the Working Class,” Institute for Policy Studies, April 6, 2026. https://ips-dc.org/attacks-on-mail-voting-are-attacks-on-the-working-class/

9. Daniel Schneider and Kristen Harknett, “It’s About Time: How Work Schedule Instability Matters for Workers, Families, and Racial Inequality,” October 16, 2019. https://shift.hks.harvard.edu/its-about-time-how-work-schedule-instability-matters-for-workers-families-and-racial-inequality/

10. Ballotpedia, “State Poll Opening and Closing Times (2026),” website accessed July 2, 2026. https://ballotpedia.org/State_Poll_Opening_and_Closing_Times_(2026)

11. Molly Roberts, “What’s up with Trump’s Mail-In Voting Executive Order?” Lawfare, Monday, June 29, 2026. https://www.lawfaremedia.org/article/what-s-up-with-trump-s-mail-in-voting-executive-order

12. Stephen Richer, “USPS Issues Proposed Mail Voting Rules Pursuant to Trump Executive Order,” Cato at Liberty blog, May 29, 2026. https://www.cato.org/blog/usps-issues-proposed-mail-voting-rules-pursuant-trump-executive-order

13. Jonathan Smith, APWU Comments on Proposed Rule: Ballot Mail for Federal Elections, June 29, 2026. https://apwu.org/wp-content/uploads/2026/06/APWU-Comments-VBM-Rulemaking.pdf

14. Dion Nissenbaum, “Federal judge blocks key pillars of Trump executive order restricting mail voting in 2026 election,” VoteBeat, June 25, 2026. https://www.votebeat.org/national/2026/06/25/trump-election-overhaul-mail-voting-executive-order-blocked-talwani-usps-dhs/

15. Molly Roberts, “What’s up with Trump’s Mail-In Voting Executive Order?” Lawfare, Monday, June 29, 2026. https://www.lawfaremedia.org/article/what-s-up-with-trump-s-mail-in-voting-executive-order

16. Smith, op. cit.

17. Ibid.

18. Ibid.

19. Ibid.

June 22, 2026

Daniel Navarrete
Director of the Division of Regulations, Legislation, and Implementation
Wage and Hour Division
U.S. Department of Labor
Room S-3502
200 Constitution Avenue NW
Washington, DC 20210

Re: Proposed Rule: Joint Employer Status Under the Fair Labor Standards Act, Family and Medical Leave Act, and Migrant and Seasonal Agricultural Worker Protection Act (RIN 1235-AA48)

Dear Mr. Navarrete,

We write to submit this comment on behalf of the Economic Policy Institute (EPI), responding to the Department of Labor’s proposed rule on Joint Employer Status Under the Fair Labor Standards Act (FLSA), Family and Medical Leave Act (FMLA), and Migrant and Seasonal Agricultural Worker Protection Act (MSPA). 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.

EPI strongly opposes the Department of Labor’s (DOL’s) proposed rulemaking and urge the agency to withdraw this rule. If implemented, we conservatively estimate this rule would cost workers roughly $1 billion annually through increases in workplace fissuring and exposure to wage theft. Further, the FLSA’s joint employer definition is also used to apply protections under FMLA, MSPA, the Providing Urgent Maternal Protections (PUMP) for Nursing Mothers Act (now part of the FLSA), and the Equal Pay Act. Under the proposed rule, workers thus would not only be at risk of losing full protections to their right to earn the minimum wage over overtime pay, but also their right to unpaid but job-protected family and medical leave, pay discrimination protections, and the right to pump breastmilk while at work. Agricultural workers, already operating in notoriously underpaid and hazardous conditions, will also find it harder to enforce or get compensation for violations of their rights to the basic pay and housing requirements for agricultural workers under MSPA. Because of the broad impacts of structural racism and sexism on labor market outcomes, women and people of color are overrepresented in low-wage jobs overall, which are particularly vulnerable to fissuring and wage theft. As a result, women workers and workers of color are likely to be disproportionately harmed if this rule is finalized.

The proposed rule would undermine the original intent of the FLSA

At its most basic, the joint employer standard simply requires that when multiple employers co-determine or share control over a workers’ terms of employment (such as pay, schedules, and job duties), each of those employers is responsible for compliance with worker protection laws. Given the realities of the modern workplace, in which employees often find themselves subject to more than one employer, workers deserve a joint employment standard under the FLSA that guarantees these basic rights and protections.

As the American Civil Liberties Union (ACLU) has argued in their joint comments, also cosigned by EPI, the NPRM contravenes the statutory definition of “employ” under the FLSA, Supreme Court precedent. This rule also shares the same substantive defects as DOL’s 2020 Final Rule, which was largely invalidated by a federal district court in New York v. Scalia, 490 F. Supp. 3d 748 (S.D.N.Y. 2020).

EPI has conducted extensive research and policy analysis on the harms to workers from weakened labor standards and weakened enforcement of those standards. There is no question that this proposed rule would weaken labor standards. As with the first Trump administration’s attempt at weakening these regulations, this rule would dramatically narrow the set of circumstances whereby a firm can be found to be a joint employer under the FLSA. The FLSA is our nation’s fundamental worker protection statute, providing wage and hour protections to the vast majority of U.S. workers. The FLSA was drafted broadly, and its definition of an employer was intended to cover most workplaces and most workers. The intention was and should remain that companies that use staffing agencies, temporary workers, or subcontractors in their business operations are held accountable for complying with the FLSA’s basic provisions, including minimum wage, overtime, and child labor protections. The proposed rule will make it nearly impossible for many workers in those types of workplaces to enforce these rights. It would also take away the ability of workers to recover unpaid wages from firms who use undercapitalized contractors in their work.

We believe it is also important to acknowledge some of the most frequently referenced critiques of a broad, protective joint employer standard, from those who would like to see that standard weakened. One argument, already present in some of the comments that the Department has received on this rule, is that this weakened standard is necessary to provide regulatory clarity for franchisee employers in particular. The International Franchise Association, for example, says this proposed rule “protects the independence of franchise small businesses.”1 However, these arguments obscure the fact that it is already large corporate franchisors who stand to benefit the most from having this “independence” protected.

Franchisee operators already bear the full responsibility for violations of the FLSA that occur on their watch, even if those violations may have been more likely to occur because of requirements or pressure exerted on them in their business agreements with the large corporate franchisors. Nothing in the FLSA’s current joint employer standard automatically labels a franchisor-franchisee relationship as a joint employment scenario. On the contrary, the longstanding joint employer standard is not one-size-fits-all, and always requires looking at multiple factors to determine how much control each entity is actually exerting on a worker. We urge the Department not to adopt a proposed rule that would continue to allow large employers to conceal their real interest—minimizing their own liability for FLSA violations—as a goal that is aligned with the best interests of small business owners and franchise operators.

The Department’s flawed economic analysis overlooks that workers will lose pay if this rule is implemented

DOL continues its misguided evaluation of the likely impacts of the proposed joint employer standard in its economic impact analysis. DOL has a responsibility to consider all relevant data in advancing this regulatory standard, but it fails to do so. The NPRM states that “the Department does not expect that there would be significant transfer effects as a consequence of the proposed rule,” explaining that “nothing in the proposed rule would reduce the wages owed to employees under the FLSA or MSPA [emphasis added].”2 

However, even if the proposed rule would not change the wages due to a worker under the FLSA or MSPA, this does not mean that the proposal will not result in transfers between employers and employees. It would, in at least two ways.

First, this rule would incentivize workplace “fissuring,” i.e., employers increasing their reliance on contractors, subcontractors, temporary help agencies, and franchises rather than hiring employees directly—a practice that suppresses workers’ wages.  The Department dismisses the idea that the rule would incentivize fissuring by essentially simply asserting that such concerns are “largely inapplicable” to this rulemaking.3  We, however, conservatively estimate that in the long run, the increase in workplace fissuring as a result of the rule would result in a transfer of at least $772.0 million from workers to employers annually. This calculation is discussed in depth below.

Second, this rule would increase losses due to wage theft by employers. The Department acknowledges that this is an issue when it states that “some workers in vertically-tiered industries may, in some cases, have more or less difficulty collecting their owed wages,” 4 but, astonishingly, dismisses this concern by stating, without evidence, that “the magnitude of this effect is unlikely to be significant.”5  We conservatively estimate that an increase in losses due to wage theft as a result of the rule will result in a transfer of at least $225.6 million from workers to employers annually. This calculation is discussed in depth below.

Putting together these two estimates—more than $772.0 million lost by workers as a result of the rule due to an increase in workplace fissuring and more than $225.6 million in losses by workers as a result of wage theft—we estimate workers will lose roughly one billion dollars ($997.6 million) annually as a result of this rule if it is finalized.

Quantifying the transfers from workers to employers due to an increase in fissuring

According to data from the Bureau of Labor Statistics’ 2023 Contingent Worker Supplement (CWS), there are 862,000 workers who work for contract firms and 945,000 workers who work for temporary help agencies.6 However, the CWS undercounts these workers. This is due in part to the fact that workers self-report what kind of firm they work for and may erroneously report that they work for the company where they are doing their work instead of for the contract firm or temporary help agency that placed them at that site. Establishment surveys—where the firm, not the worker, does the reporting—get around this problem. High-quality establishment data on employment in contract firms do not exist to our knowledge, but there are excellent establishment data on employment in temporary help agencies from the Bureau of Labor Statistics’ Current Establishment Survey (CES). These data show that there were 2.50 million workers in temporary help agencies in 2025,7 well over double (2.64 times) what is reported in the latest CWS. Adjusting the number of contract workers by the same multiple (2.64) results in an adjusted estimate of the number of contract workers of 862,000 * 2.64 = 2.28 million.

It is important to note that we believe this estimate still undercounts contract workers, because the CWS includes only one very specific type of contract worker in its count of workers employed by contract firms—workers who are usually assigned to only one client and usually work at the client’s worksite. That excludes the many contract workers who work for multiple clients (e.g., janitorial workers or IT consultants) or offsite (e.g., call center workers or industrial laundry workers). We do not attempt to quantify this undercount.

Another important form of fissuring in the workplace is the increasing reliance on franchising models. Data from the U.S. Census Bureau’s 2017 Economic Census Franchise Statistics Report show that franchise employment in 2017 in key sectors where franchising is common was 9.59 million.8  Since 2017 is the latest year these data are available, we inflate the value by the growth rate in overall payroll employment between 2017 and 2025, 8.1%, from the Current Employment Statistics establishment survey of the Bureau of Labor Statistics.9  This results in an estimated level of franchise employment for 2025 of 10.36 million.

Putting this all together, we conservatively estimate that in 2025, there were a total of 15.14 million employees working in “fissured establishments”—working for temporary help agencies (2.50 million), working for contract firms (2.28 million), or working for franchises (10.36 million). It is important to note the degree to which this estimate of the fissured workplace is likely an undercount. David Weil estimated that in 2017, 18.9 percent of private-sector production and nonsupervisory workers—20.8 million workers in 2025—were in highly fissured industries, and that if additional fissured workers in occupations and in industries with mixed use of practices were included, that share could easily double.10

Because the rule would mean that employers would be able to avoid liability for FLSA violations for many workers in fissured establishments while still substantially controlling the wages and working conditions of those workers, companies will be incentivized to restructure and outsource parts of their business. Research shows that the wage losses associated with this kind of domestic outsourcing are substantial, on the order of 5% long-run earnings losses.11 Thus the rule will result in a substantial transfer away from workers whose firms decide, as a result of the rule, to outsource the work that they do.

CES data show that the average weekly earnings of production and nonsupervisory workers in temporary help services in 2025 was $932.12 A 5% penalty (noted above) for working in a fissured workplace implies that these workers would be earning $981 if they were directly hired, a difference of $49 per week. Combined with our estimate of 15.14 million employees working for fissured establishments, we find that every percent increase in fissuring as a result of the rule would, in the long run, lead to a wage loss of $386.0 million annually.13 That means that an increase in domestic outsourcing of just 2 percent as a result of the rule—an implausibly conservative increase considering employers would newly be able to avoid liability for FLSA violations while still substantially controlling the wages and working conditions of domestically outsourced workers—would lead to a transfer of $772.0 million annually from workers to employers. Further, it is important to note that using the broader estimate, described above, of 20.8 million private-sector production and nonsupervisory workers in the fissured workplace, that number would be $1.1 billion.

Quantifying the transfers from workers to employers due to an increase in wage theft

Wage theft—the practice of employers failing to pay workers the full wages to which they are legally entitled—is a widespread and deeply rooted problem that directly harms millions of U.S. workers each year. Employers refusing to pay promised wages, paying less than legally mandated minimums, failing to pay for all hours worked, or not paying overtime premiums deprives working people of billions of dollars annually. It also leaves hundreds of thousands of affected workers and their families in poverty.14 Wage theft does not just harm the workers and families who directly suffer exploitation; it also weakens the bargaining power of workers more broadly and puts downward pressure on hourly wages in affected industries and occupations. For many low-income families who suffer wage theft, the resulting loss of income forces them to rely more heavily on public assistance programs, unduly straining safety net programs and hamstringing efforts to reduce poverty.

In 2008, Bernhardt et al. surveyed front-line workers in low-wage industries in the cities of Chicago, Los Angeles, and New York and found that two-thirds (68 percent) of these workers experienced at least one pay-related violation in any given week.15 The researchers estimated that the average cost to these workers over a year was $2,634 out of a total earnings of $17,616—15.0 percent of their wages. This adds up to a total of nearly $3 billion annually stolen across all forms of wage theft among these workers in 2008. Generalizing these three-city, 2008 results to the nationwide 2025 workforce, we estimate that low-wage workers in the U.S. lost $52.5 billion to all forms of wage theft in 2025.16

It is worth noting that though the Bernhardt et al. study is somewhat dated, more recent studies show that labor violations remain so prevalent that it is likely that simply extrapolating from the Bernhardt et al. study, as we have, will generate conservative numbers.  For example, a 2024 study out of the Shift Project at Harvard Kennedy School found that nearly all (91%) hourly service sector workers in California experienced at least one labor violation in the prior year.17

The proposed rule would increase losses due to wage theft by employers in at least three ways. Each of these impacts will be particularly acute in industries in which there is high reliance on subcontracting, temporary work, and other alternative work arrangements, where there is already a disproportionate occurrence of wage theft.18

First, the proposal would severely limit the ability of millions of workers to get justice when they are victims of wage theft. By limiting workers’ ability to recover wages from firms that contractually have the right to act with respect to the terms and conditions of employment, DOL is depriving workers of long-held rights to recover unpaid wages from their employers.

Second, there will be a reduction in wage theft deterrence caused by the reduction, as a result of the rule, of workers’ ability to recover wages. This reduction in wage theft deterrence will likely lead to an increase in wage theft.

Third, by allowing firms that hire contractors to avoid legal liability for wages, the rule would give these firms greater incentive to award contracts to undercapitalized firms that are more likely to have low bids on the basis of not paying their workers what they are owed. And, absent the legal liability stemming from being a joint employer, if the contractor goes out of business, the lead business is not liable for the lost wages of the workers. In other words, this rule would increase the incentive for firms to seek out undercapitalized contractors who will provide lower bids to the companies that use them—bids that are able to be so low because the contractors plan to steal from their workers (by underpaying them or not paying them at all).19

As described above, an estimated $52.5 billion was lost by low-wage workers to all forms of wage theft in 2025. We use several sources of data to estimate how much of that $52.5 billion was lost by workers in fissured establishments. As noted above, we conservatively estimate that in 2025, there were a total of 15.14 million employees working in “fissured establishments”—working for temporary help agencies (2.50 million), working for contract firms (2.28 million), or working for franchises (10.36 million).

To determine how many of these 15.14 million workers are low-wage, we turn to CWS microdata, which allow us to calculate the share of workers in contract firms and temporary help services who are low wage workers.  Unfortunately, microdata from the most recent (2023) CWS survey have not yet been released, so we use microdata from the 2017 CWS survey. We find that the share of workers in contract firms or in temporary help services who are low-wage—defined as earning $12 per hour or less in 2017—is 36.3 percent. Given wage growth between 2017 and 2025, $12 in 2017 was roughly equivalent to $17 in 2025.20 

Franchise workers are not identified in the CWS, so we simply assume that the share of workers in franchise firms who are low-wage is the same as the share of workers who are low-wage in contract firms and temporary help services. Multiplying 36.3% by our estimate of 15.14 million total workers in fissured establishments, we estimate that there are 5.5 million low-wage workers in fissured establishments.

There were 25.6 million workers who made less than $17 an hour in 2025,21 which means that 21.5 percent (5.5 million/25.6 million) of low-wage workers are in fissured establishments. Assuming that the incidence of wage theft among low-wage workers is no higher in fissured establishments than in traditional establishments (an extremely conservative assumption), we can simply multiply this 21.5 percent by the total amount of wage theft from low-wage workers—$52.5 billion—to estimate the amount of wage theft in fissured establishments. This comes out to $11.28 billion.

Annual wage theft of $11.28 billion in fissured establishments means that every percent increase in losses due to wage theft would lead to an aggregate transfer from workers to employers of $112.8 million annually. This means that an increase in losses due to wage theft of just 2 percent as a result of the rule—an implausibly conservative increase considering many former joint employers would newly be able to avoid liability for FLSA violations—would lead to an aggregate transfer from workers to employers every year of $225.6 million. Further, it is important to note that using the broader estimate described above of 20.8 million private-sector production and nonsupervisory workers in the fissured workplace, that number would be $309.8 million.

Conclusion

DOL’s proposed rule undermines the original intent of our nation’s fundamental worker protection laws and, if implemented, its impact on working people will be negative and significant. The proposed rule would incentivize the further “fissuring” of the workplace, putting strong downward pressure on wages, and it would make it nearly impossible for millions of workers to get justice when they are the victims of wage theft. Conservatively, we estimate that, if implemented, this rule would cost workers just under $1.0 billion annually—more than $772.0 million due to wage suppression from an increase in workplace fissuring and more than $225.6 million from an increase in wage losses due to wage theft by employers. We urge DOL to abandon this flawed rulemaking and ensure a meaningful joint employer standard under the FLSA, our nation’s fundamental worker protection law.

Sincerely,

Heidi Shierholz, Ph.D.
President
Economic Policy Institute

Samantha Sanders
Director of Government Affairs & Advocacy
Economic Policy Institute

Endnotes

1. International Franchise Association (IFA). 2026. “IFA Praises Trump Administration Joint Employer Rule” (press release). April 22, 2016.

2. 91 Fed. Reg. 21909

3. 91 Fed. Reg. 21909

4. 91 Fed. Reg. 21909

5. 91 Fed. Reg. 21910

6. Bureau of Labor Statistics, “Table 5. Employed workers with alternative and traditional work arrangements on sole or main job by selected demographic characteristics, July 2023,” Contingent and Alternative Employment Arrangements, November 2024.

7. U.S. Bureau of Labor Statistics, All Employees, Temporary Help Services [TEMPHELPS], retrieved from FRED, Federal Reserve Bank of St. Louis. Accessed June 2026 at https://fred.stlouisfed.org/series/TEMPHELPS. 

8. U.S. Census Bureau, “Franchising in America: Key Data from the 2017 Economic Census,” September 2021.

9. U.S. Bureau of Labor Statistics, All Employees, Total Nonfarm [PAYEMS], retrieved from FRED, Federal Reserve Bank of St. Louis. Accessed June 2026 at https://fred.stlouisfed.org/series/PAYEMS.

10. David Weil, “Understanding the Present and Future of Work in the Fissured Workplace Context,” Working Paper, Brandeis University, May 2019.

11. Dorn, D., Schmieder, J. F., Spletzer, J. R. (2018). Domestic Outsourcing in the United States. Chief Evaluation Office, U.S. Department of Labor; Deborah Goldschmidt and Johannes F. Schmieder, “The Rise of Domestic Outsourcing and the Evolution of the German Wage Structure,” Quarterly Journal of Economics 132, no. 3 (August 2017): 1165–1217; Arindrajit Dube and Ethan Kaplan, “Does Outsourcing Reduce Wages in the Low-Wage Service Occupations? Evidence from Janitors and Guards,” ILR Review 63, no. 2 (January 2010): 287–306.

12. Bureau of Labor Statistics, Current Employment Statistics (BLS-CES). Table B-8, Average hourly and weekly earnings of production and nonsupervisory employees on private nonfarm payrolls by industry sector, seasonally adjusted. Various years. Accessed June 2026 at https://www.bls.gov/webapps/legacy/cesbtab8.htm.

13. $386.0 million = 15.14 million * $49 * 52 weeks in a year * 1%.

14. Margaret Poydock and Jiayi (Sonia) Zhang, More than $1.5 billion in stolen wages recovered for workers between 2021 and 2023, Economic Policy Institute, December 2024; David Cooper and Teresa Kroeger, Employers Steal Billions from Workers’ Paychecks Each Year: Survey Data Show Millions of Workers Are Paid Less Than the Minimum Wage, at Significant Cost to Taxpayers and State Economies, Economic Policy Institute, May 2017.

15. Annette Bernhardt et al., Broken Laws, Unprotected Workers: Violations of Employment and Labor Laws in America’s Cities, 2009, Center for Urban Economic Development, National Employment Law Project, and UCLA Institute for Research on Labor and Employment, 2009.

16. Generalizing the three-city, 2008 results to the nationwide 2025 workforce required several adjustments. The low-wage workforce in the Bernhardt et al. study represented 15.1% of all workers in those cities, and 68% of those workers experienced at least one pay-related violation in the prior week. This implies that at least 15.1%*68% = 10.3% of all workers experienced wage theft in a given week. Data from the BLS Current Employment Statistics (CES) survey shows there were158.5 million nonfarm payroll employees in 2025 nationwide. Applying the 10.3% estimate to that workforce yield 16.3 million workers, meaning that at least 16.3 million workers nationwide likely experienced wage theft in any given week in 2025. Bernhardt et al. found that workers who experienced wage theft lost, on average, 15% of their weekly earnings. Using the BLS Current Population Survey (CPS), we find that the lowest-paid 15.1% of workers who are 18 years old or older and worked at least five hours per week —a conservative proxy for the population surveyed in Bernhardt et al.—had median weekly earnings of $352 in 2025. Assuming 15% losses due to wage theft, the earnings of workers experiencing wage theft would have been $414 if wage theft hadn’t occurred, an average loss of $62. Multiplying the estimated16.3 million workers experiencing wage theft by the average loss of $62, we find that the total amount lost by low wage workers to wage theft in a given week is $1.01 billion. Annualized, this amounts to $52.5 billion in wages stolen from low-wage workers each year.

17. Daniel Schneider, Elizabeth Kuhlman, Kristen Harknett, and David Weil. 2024. Compliance and the Complaint Gap: Labor Standards Violations in the California Service Sector. The Shift Project at Harvard Kennedy School, May 2024.

18. Annette Bernhardt et al., Broken Laws, Unprotected Workers: Violations of Employment and Labor Laws in America’s Cities, 2009, Center for Urban Economic Development, National Employment Law Project, and UCLA Institute for Research on Labor and Employment, 2009.

19. This is an argument made by Judge Easterbrook in Reyes v. Remington Hybrid Seed Co. 495 F.3d 403 (7th Cir. 2007). In that decision, Easterbrook notes, “If Zarate [the labor broker in the case] had been solvent, Remington [the lead business in the case] would have to offer him enough that he could pay all the workers’ wages (including the minimum wage and any overtime premium), cover the costs of fringe benefits such as housing, and still be able to make a profit. But when a contractor has no business or personal wealth at risk, he may be tempted to stiff the workers (as Zarate did) and then treating the principal firm as a separate employer is essential to ensure that the workers’ rights are honored.”

20. $12 was 66.1% of the median wage in 2017, and 66.1% of the median wage in 2025 was $16.97.  Economic Policy Institute, State of Working America Data Library, “Hourly wage percentiles – Nominal hourly wage,” 2026.

21. Low-Wage Workforce Tracker, Economic Policy Institute, January 2026.

June 1, 2026

The Honorable Daniel Aronowitz
Assistant Secretary
Employee Benefits Security Administration
U.S. Department of Labor
200 Constitution Avenue NW
Washington, DC 20210

Re: Fiduciary Duties in Selecting Designated Investment Alternatives, RIN 1210-AC38

Dear Assistant Secretary Aronowitz:

I submit these comments on behalf of the Economic Policy Institute (EPI) on the Department of Labor’s (DOL) Notice of Proposed Rulemaking on Fiduciary Duties in Selecting Designated Investment Alternatives.

EPI is a nonprofit, nonpartisan think tank that has worked for 40 years to center working families in economic policy discussions.

EPI strongly opposes the Employee Benefits Security Administration’s (EBSA’s) proposed rule on Fiduciary Duties in Selecting Designated Investment Alternatives. The rule would gut protections for retirement savers in 401(k) and other participant-directed plans covered by the Employee Retirement Income Security Act of 1974 (ERISA). In addition to our comments here, EPI has signed onto a joint comment letter with other consumer, retiree, and worker advocacy organizations expressing concern that the rule would add complexity, raise fees, and return us to an era when retirement plan participants had little protection from poor investment choices. This letter will expand on that letter to critique some of the economic arguments made in support of the proposed rule.

This letter will focus on risk, questioning EBSA’s claim that loosening protections for retirement savers will maximize risk-adjusted returns by providing access to alternative investments.

• EBSA attempts to redefine fiduciaries’ duty of prudence in selecting investment options as “maximizing risk-adjusted returns net of fees” no matter the risk rather than balancing the goals of maximizing returns and minimizing risk. (Unless otherwise noted, quotations are from the proposed rule.)
• EBSA ignores the fact that retirement savers face a worse principal-agent problem than defined benefit pension funds, since investment options are chosen by plan fiduciaries but participants bear the risk of losses.
• EBSA ignores the fact that even if pension funds that invest in alternative assets earn an illiquidity and risk premium (a big “if”), and even if individual retirement savers retire at the time they planned to (another big “if), individual retirement savers with limited investment horizons face greater timing risk and should be more risk averse.
• EBSA assumes that any asset class that does not move in sync with stock and bond markets adds useful diversification even if it is a speculative asset with a zero expected real return that only adds volatility.
• EBSA’s case for asset class diversification rests on modern portfolio theory, which assumes investors are well-informed and risk-averse, which is not a realistic description of markets for alternative assets.
• EBSA focuses narrowly on asset class diversification, ignoring the fact that adding alternative assets can concentrate rather than spread risk if these assets cannot be indexed.

Plan fiduciaries must adhere to a duty of prudence and a duty of loyalty.

Under ERISA, fiduciaries may be sued by participants or the Department of Labor (DOL) for breach of these duties when selecting investment options for participant-directed plans. Fiduciaries are required to have or obtain the expertise and information to make substantively sound decisions—going through the motions or acting with good intentions is not enough.

With minor exceptions, such as limits on employer stock, ERISA does not specify what types of investments may be included in 401(k) and similar plans. Nevertheless, plan fiduciaries, using their good judgement and leery of lawsuits, have generally avoided alternative investments that cannot be marketed to small and unsophisticated (nonaccredited) investors in other contexts. Instead, fiduciaries have increasingly moved toward low-cost index funds and similar broadly diversified and publicly traded investments that are recognized as appropriate options for retirement savers.

Lawsuits on behalf of plan participants who suffer losses from fiduciaries’ failure to uphold their duties are a critical enforcement tool. According to AARP, “Courts have repeatedly recognized that these cases play a meaningful role in improving plan governance, reducing excessive fees, and protecting participants’ long-term retirement security, without burdening employers that comply with the law.”1

The stated purpose of the proposed rule is to expand access to alternative investments.

EBSA claims that the threat of lawsuits harms retirement savers by limiting investment options available to plan participants. EBSA claims that “regulatory burdens and litigation risk…interfere with the ability of American workers to achieve…the competitive returns and asset diversification necessary to secure a dignified and comfortable retirement.”

The proposed rule includes a list of alternative assets that President Trump earlier mentioned in an executive order, including private equity, private credit, digital assets (“crypto”), and commodities.

EBSA notes that under ERISA these assets are not explicitly prohibited from being included among investment options in participant-directed plans. However, EBSA warns that plan fiduciaries may be excluding more complex investment options “not necessarily in response to a prudent assessment of whether the features in those investments are best suited to the needs of plan participants and beneficiaries, but rather because of the risk of litigation if plan fiduciaries depart from more traditional investments in favor of more creative or novel options.” It provides no evidence for that assertion and does not rebut the valid presumption that this could be the framework operating as intended: fiduciaries have made prudent assessments, found these assets wanting, and decided against them— recognizing that doing otherwise would appropriately expose them to litigation risk.

EBSA reframes the goal as maximizing risk-adjusted returns” no matter the risk rather than balancing risk and return.

According to EBSA, ERISA gives fiduciaries “maximum discretion and flexibility” to determine which investment options “offer the opportunity for participants to maximize risk-adjusted returns on their retirement assets net of fees.”

However, the goal of ERISA, as the name makes clear, is retirement income security, not “maximizing risk-adjusted returns net of fees.” Under ERISA, fiduciaries are required to select a diversified menu of investment options that minimize the risk of large losses for retirement savers.2 Such losses can be caused by sharp downturns or by cumulative underperformance over many years. Retirement savers will face both types of risk if they invest directly or indirectly in the alternative assets listed in the proposed rule.

Leverage is the most obvious way to maximize risk-adjusted returns, but also the most obvious way to amplify the risk of large losses. Reputable financial advisors do not advise retirement savers to maximize risk and return through leverage, even in the case of early career workers who have more time to adjust their contributions to the plan if their risky strategy fails.

EBSA ignores the subject of leverage entirely in the proposed rule, as well as any discussion of the appropriate level of portfolio risk for retirement savers (except to claim that asset diversification reduces it). Implicitly, then, EBSA’s self-described “asset-neutral” approach is also neutral with respect to the amount of risk retirement savers should face, focusing only on maximizing returns for any level of risk. This is especially dangerous given the widespread but mistaken belief that investment returns will average out over long investment horizons.3

Encouraging risky investments harms not only unsophisticated retirement savers, but also taxpayers who subsidize accounts intended to help ordinary workers save for retirement.

Though certain risky investment options may benefit wealthy participants with a high risk tolerance, the tax advantage is intended to promote retirement savings, and retirement security, for ordinary workers, not amplify wealth inequality.

Regulatory agencies have repeatedly warned of alternative asset risks.

DOL and other regulators previously acknowledged the risks associated with private market investments, including opacity, lack of regulatory oversight, complexity, illiquidity and high fees. In a June 3, 2020, letter to private equity managers, DOL noted that private equity investments had longer time horizons, higher fees, no easily observed market value, and were subject to different regulatory requirements and oversight than publicly traded securities.4

The letter suggested that plan fiduciaries might want to limit private equity investments to a specified percentage of a fund, have the investments independently valued according to agreed-upon financial standards, and require additional disclosures to meet the plan’s ERISA obligations to report information about the current value of the plan’s investments. These suggestions were generally ignored in the proposed rule.

After the Securities and Exchange Commission (SEC) issued a risk alert on June 23, 2020,5 warning that private equity and hedge fund investors may have paid more in fees and expenses than they should have and may not have been informed of conflicts of interest,

DOL issued a supplemental statement on December 21, 2021,6 citing the SEC warning and stakeholder comments challenging the earlier letter’s uncritical acceptance of some industry talking points, notably the claim that private equity could “offer plan participants who have longer investment horizons an equities-based investment choice that may enhance retirement outcomes when compared to investment choices containing only publicly traded securities.” The DOL statement also noted that while some fiduciaries have experience evaluating private equity investments for defined benefit pensions, many fiduciaries of small individual account plans do not.

The DOL statement was rescinded in response to President Trump’s executive order of August 7, 2025,7 without addressing any of the issues raised about the advisability of including private market investments among 401(k) plan options.

DOL and other regulators have also previously been highly skeptical of crypto as an investment, let alone one offered to retirement savers. On May 11, 2021, the SEC issued a staff statement warning that Bitcoin and Bitcoin futures were highly speculative investments.8 On January 10, 2024, SEC Chair Gary Gensler went further, describing Bitcoin as “primarily a speculative, volatile asset that’s also used for illicit activity including ransomware, money laundering, sanction evasion, and terrorist financing.”9 The Federal Reserve and other bank regulatory agencies also issued multiple warnings about crypto assets.10

Citing the SEC warnings, DOL issued guidance on March 10, 2022, advising 401(k) plan fiduciaries to exercise “extreme care” before adding cryptocurrencies to plan options, noting that they were difficult to valuate, even by experts, and posed custodial and recordkeeping concerns.11

The Federal Reserve and other bank regulators rescinded their guidance on April 24, 2025, and DOL followed suit on May 28, 2025, without addressing any of the issues raised about crypto risks. Beginning in 2025, the SEC also adopted a more accommodating stance toward crypto, for example, asserting that meme coins were not subject to federal securities laws12 and dismissing an enforcement action against Coinbase.13

The SEC prevents firms from marketing private assets to small (“retail”) investors for good reasons.

Laws including the Investment Company Act and Investment Advisers Act, both enacted in 1940, give the SEC the authority to regulate securities marketed to retail investors.14 In addition to requiring consistent valuations and disclosures, these laws—and regulations and guidance based on them—limit the use of leverage and guard against potential conflicts of interest.15 The sale of private funds that do not meet these requirements is generally limited to sophisticated “accredited” investors or “qualified purchasers.”

These laws are intended to protect investors who are not equipped to assess the value or risk of private market investments. During the Biden administration, the SEC proposed rules that recognized that even accredited investors in private funds were not provided with sufficient information or protection from conflicts of interest.16 The rules were overturned by the Fifth Circuit Court of Appeals based on the assumption that investors in these funds were sophisticated and able to bear losses, not that they had adequate information and protection.17

Are risk, illiquidity, and complexity opportunities to earn higher returns?

EBSA acknowledges some of the obvious disadvantages of alternative investments, but frames these as opportunities to earn a premium for accepting these risks:

Alternative asset investments are often less liquid than the publicly traded stock and bond funds that are held by funds that plan fiduciaries often make available to plan participants. Illiquid investments generally offer an illiquidity premium to investors who are willing to hold their investment, for some time, without selling it for cash. Many retirement savers, particularly younger workers, have long investment time horizons until retirement and, therefore, fit the profile of an investor who can benefit from a liquidity premium.

EBSA elsewhere alludes to “obstacles that cause a relatively higher risk premium when compared to traditional investments, such as illiquidity or information asymmetry,” but suggests that an expanded market tailored to the needs of retirement savers could reduce liquidity and valuation risks:

As this market matures a new equilibrium should be reached where there is a larger pool of viable and vetted investments, expanded by alternative assets, for asset managers to o􀆯er to plan sponsors. The tradeoff for this increased market penetration is a reduction in illiquidity premium. As the risks associated with investment in alternative assets falls, so too does the risk premium investors in the assets will enjoy.

The idea that retirement savers could knowledgeably accept such tradeoffs ignores the fundamental problem that information asymmetry, complexity, and lack of regulatory oversight make reliable ex ante valuation of private market assets impossible.18 Expanding the market, whether or not it improves liquidity, does not address this fundamental problem.

Private market investments should be niche products for sophisticated investors.

The accurate valuation of public assets is made possible by the mandatory disclosure of relevant information and market signals from daily trading. A key feature of public markets is two-sided competition, wherein knowledgeable buyers and sellers eliminate biased prices so that even unskilled and uninformed investors can get a fair shake.19 These sources of information are generally unavailable to private market investors, who in most cases must also contend with complex structures, strategies, and incentives.

Private markets are characterized by asymmetric information, so outside investors cannot assume that investment opportunities are fairly priced. Institutional investors instead rely on private fund managers’ past performance, reputational risk, and performance incentives in deciding whether to enter into limited partnerships with fund managers, known as general partners. Institutional investors with clout can also demand access to information that may not be provided to all limited partners and would not be made available to retirement savers. However, general partners’ incentives are blunted and distorted by their ability to earn millions in fees, and often in related party transactions, even when funds underperform or sustain losses, a problem exacerbated by the favorable tax treatment of fund managers’ share of investment returns, known as carried interest. Past performance, meanwhile, has been found to be a weak predictor of future performance.

In a comment submitted about the proposed rule, former investment banker Jeffrey Hooke and business school professor Michael Imerman estimated that private market investments should earn a premium of 200-500 basis points over their public counterparts to compensate for illiquidity, leverage and opaque financial reporting. Far from earning such a premium, Hooke and Imerman estimate that over 95% of state pension plans investing in private assets fail to beat a 60-40 stock-bond benchmark over long periods.20 Even if some institutional investors with superior clout and expertise can expect higher risk-adjusted returns by investing in private market assets, the wide dispersion in fund performance suggests that other institutional investors are not assured of benefiting from these investments and that retirement savers would fare even worse.21

The expected real return on purely speculative assets is zero.

With private market investments, the challenge lies in gauging the expected risk and return of underlying assets. With most crypto and other speculative assets, there is no expectation of profit unless the buyer has better information than the seller. At best, these assets are “digital gold,” used as a store of value, a means of exchange, or a hedge against inflation, but with added risks associated with custody, safekeeping, and regulatory uncertainty.22

As the SEC acknowledged last year, some crypto assets can be compared to physical collectibles like rare tulip bulbs, baseball cards, and Beanie Babies because they are nonproductive investments.23 Under the Economic Recovery Tax Act of 1981, most physical collectibles are banned from tax-favored retirement plans because they “do not contribute to productive capital formation.”24 Though digital collectibles, unlike physical collectibles, are not currently banned from investment options offered to participants in tax-favored retirement plans, it is not clear why they should be exempted.25

Similar to much crypto, commodity futures are largely speculative and regulated by the Commodities Future Trading Commission (CFTC), though investors may be compensated for hedging the risk of producers or consumers. The New York Times recently described how the CFTC’s enforcement capacity has been hollowed out and the agency has become captive of crypto and prediction markets.26

A market for speculative assets exists for the same reason there are casinos, racecourses, and prediction markets: A subset of the population, rather than being risk averse, likes to gamble. However, gambling does not belong in tax-subsidized accounts intended to promote retirement income security.

Assessing the performance of private funds poses serious challenges.

The debate around whether investing in private market assets is worth the high fees, risk, and illiquidity is complicated by a lack of consistent disclosure requirements. As documented by Oxford University professor Ludovic Phalippou and others, private equity general partners, when marketing themselves to pension funds and other potential investors, cite irrelevant or misleading statistics, sometimes manipulating the timing of valuations or excluding funds that have been committed but not yet invested to inflate reported returns.27

Given the subjectivity of internal valuations and evidence that they are misleading and manipulated, limited partners’ return on investment can only be known once assets have been sold and the proceeds distributed. This happens too infrequently to be useful in making investment decisions, a problem exacerbated by the fact that private fund managers can prevent limited partners from fully exiting.28

Even after the return on investment is known to limited partners, it is difficult to know whether the return was sufficient to compensate for risk and illiquidity, since the degree of leverage and other risk factors is generally unknown to outside investors.

The available evidence does not show that alternative assets improve risk-adjusted returns. Because performance metrics reported by private equity and other alternative assets are unreliable, researchers have looked at whether institutional investor portfolios that include these investments have outperformed benchmarks composed of broad stock and bond indices. Many found that they did not, especially in the years since the 2008 financial crisis.29

For example, a 2022 report from the Center for Retirement Research at Boston College found that public pension funds that invested more in alternative investments did not have higher returns, though the investments may have served to dampen reported volatility.30 Similarly, researchers at the Canada Pension Plan Investment Board found that while private equity appeared to outperform stocks before the financial crisis, it did not do so on a risk-adjusted basis.31

A more positive study published by the Georgetown Center for Retirement Initiatives found that 401(k) participants would have seen slightly higher returns over a 20-year period if target date funds had included private equity and other alternative investments.32

However, even this industry-funded study showed that large-cap U.S. stocks outperformed private equity in the decade after 2011. The study relied on a proprietary database of pension fund returns that is subject to major revisions and may not include a representative sample of pension funds.

Perhaps the most telling indicator of private funds’ mediocre performance is the industry’s resistance to providing comparable metrics even to their own investors. In 2024, after the SEC attempted to standardize information about fees and performance provided to limited partners in private funds,33 the industry challenged the rule before the Fifth Circuit Court of Appeals, which sided with the industry on the basis that access to the funds was generally limited to “some of the most sophisticated and wealthiest investors.”34 Of course, this will no longer be true if private assets are marketed to retirement savers.

The proposed “safe harbor” does not protect retirement savers—or fiduciaries.

At the heart of the proposed rule is what EBSA describes as a process-based “safe harbor” that would shield fiduciaries who adhere to the process from liability even if the investment options they select for the plan are poor choices. EBSA lists six factors that fiduciaries should consider when selecting investment options: performance, fees, liquidity, valuation, benchmarking, and complexity.

In EBSA’s view, fiduciaries who follow the six-step process should be granted a “presumption of prudence” and given “the discretion and flexibility to determine when designated investment alternatives, including those that contain alternative investments, offer the opportunity for participants to maximize risk-adjusted returns on their retirement assets net of fees.”

This check-the-box process does not give fiduciaries the tools they need to assess opaque, complex, illiquid, and largely unregulated investments. The proposal also does not address disclosures to plan participants, who would be even less prepared to make informed decisions.

The types of alternative investments listed in the proposal, by their nature, cannot be reliably evaluated according to the criteria in the process-based rule, even assuming fiduciaries have the skill and experience to understand complex structures and incentives.

Without additional disclosure requirements and regulatory oversight, for example, it is impossible to reliably assess risk and liquidity in private assets in order to choose a comparable benchmark.

ERISA lawyers warn that the regulatory “safe harbor” may give fiduciaries a false sense of security without shielding them from lawsuits.35 The proposed rule assumes that courts will defer to the agency in granting a “presumption of prudence” to fiduciaries, even though no such presumption exists in the statute and the U.S. Supreme Court recently ruled in Loper Bright that courts should exercise independent judgement in interpreting laws rather than deferring to agency interpretations.

EBSA assumes that diversifying across asset classes reduces risk.

EBSA emphasizes that ERISA does not restrict plan fiduciaries from considering any type of asset class, suggesting that all types should be made available to retirement savers. This is akin to arguing that since USDA dietary guidelines do not explicitly ban junk food, it should be part of every diet.

EBSA cites modern portfolio theory to argue that “the optimal constrained portfolio will have lower risk-adjusted returns than that of an unconstrained portfolio.” Modern portfolio theory is a highly abstract model that relies on unrealistic assumptions, including perfect information and perfectly liquid markets, that clearly do not describe private markets.

Though markets for crypto and other speculative assets can be highly liquid and do not depend on access to information, modern portfolio theory also assumes that investors are risk averse, and risk-averse investors do not engage in pure speculation.

Modern portfolio theory can be a useful approximation of reality, for example, in support of passive investment in publicly traded securities. But it cannot preclude the existence of overpriced assets designed to lure naïve investors into private markets where information is asymmetric and the “smart money” has no way to capitalize on better information through short selling.

Diversification across asset classes does not necessarily reduce risk, especially if the asset that is added is itself not diversified.

Diversification across asset classes can be a valid reason to expand the range of available investment options to retirement savers. However, whether an asset class will improve risk-adjusted returns depends on net returns, volatility, and correlation with stocks and other portfolio assets.

Alternative investments are often touted for their supposed low volatility and low correlation with stocks. But as Morningstar and others have pointed out, this reflects infrequent valuations and should not be mistaken for low risk: “With low disclosure and transparency, frequent use of leverage, and valuations that are both lower in scale and frequency than public markets, [private investments] should be considered one of the riskiest asset classes in an investor’s portfolio, despite often being sold as having lower risk profiles.”36

Private assets tend to be correlated with their public counterparts: private equity with stocks, private credit with bonds, etc. Speculative assets such as crypto may be less correlated with traditional investments, but the volatility in returns is essentially noise— adding risk without a risk premium since investors in these assets are not risk averse. EBSA’s narrow focus on asset class diversification ignores the fact that adding alternative assets can concentrate rather than spread risk if the assets themselves are not diversified.

A major concern is that indexing is generally not possible with private assets. A target date fund composed of broad market indices is more diversified than one that includes a 15% private equity fund with holdings in 10 companies selected and overseen by the same general partners.

There are important differences between pension funds and individual retirement savers.

The proposed rule implements a section of President Trump’s August 7, 2025 Executive Order, “Democratizing Access to Alternative Assets for 401(k) Investors,” which argues that that since most defined benefit (DB) pension funds invest in alternatives, participants in 401(k) and other defined contribution (DC) plans are being disadvantaged.37

This ignores key differences between DB pensions and DC plans, notably the fact that fiduciaries who select DB pension investments are employed by the party that bears the risk (the plan sponsor). In contrast, DC plan sponsors and their fiduciary advisors choose the investment options but the risk of losses falls on retirement savers.

Illiquid investments can trip up retirement savers who need to tap their savings earlier than planned due to unanticipated health shocks, job loss, or caregiving responsibilities. Firms marketing target date and asset allocation funds with limited private asset “sleeves” assure retirement savers that between the more liquid assets in the fund and incoming contributions, cashing out early will not be a problem. However, this assumes a stable market—no panic selling—and that investors who cash out will not be saddled with high fees with little to show for it.

Less often noted than the illiquidity issue is the fact that timing (or “sequence-of-returns”) risk should make retirement savers more risk averse than pension funds and other institutional investors with indefinite investment horizons even if they do not cash out early.38 Whereas pension funds and university endowments continuously buy and sell assets on behalf of overlapping generations of beneficiaries, retirement savers are more sensitive to poor investment performance over limited investment horizons, especially once workers have built up savings and are approaching retirement.

Retirement savers will be defaulted to high-cost and risky investments.

EBSA suggests that both direct and investments in alternative assets should be permitted, raising the possibility that retirement savers—including those close to retirement—could put all their retirement savings in crypto and other risky assets. However, EBSA suggests that “a more likely scenario…is that these alternatives would be included as one part of a menu option,” citing as examples target date funds and asset allocation funds with annuity or private equity components similar to those the industry has begun marketing to retirement savers. Target date and balanced asset allocation funds can be “qualified default investment alternatives” (QDIAs) into which participants can be defaulted unless they opt out.

The idea that alternative investments could be embedded in target date and similar funds is not reassuring. Retirement savers who are defaulted into QDIAs are generally unsophisticated investors who will need to rely on their investments to cover living expenses in retirement and have been led to believe that these defaults are relatively safe. DOL’s website describes default investments as investments “that generally minimize the risk of large losses and provide long term growth.”39

If the proposed rule takes effect, many workers could unknowingly invest 15% or more of their retirement savings in assets that DOL has previously characterized as high-cost, complex, illiquid, and difficult to evaluate, reversing what has been a welcome shift to low-cost passive investments.

401(k) investment options have improved in recent years with the widespread adoption of target date and balanced funds composed of low-cost stock and bond indices and similar broadly diversified passive investments. This has enabled retirement savers to lower costs, automatically rebalance portfolio allocations, adjust portfolio risk as workers approach retirement, and maximize diversification across publicly listed securities.

The rule would harm not only retirement savers, but also the broader economy.

The aggregate value of largely unregulated private funds, including both private equity and private credit, now approaches that of regulated public funds.40 While it is highly concerning that unregulated private markets are growing at the expense of public ones, the solution is extending disclosure requirements and other investor protections to private markets, not increasing the size of unregulated markets that expose investors and other economic actors to exploitation and excessive risk.

Private equity has often been a destructive force in the economy. It has a reputation for loading companies up with debt, stripping them of assets, and often driving them into bankruptcy, leaving workers, suppliers, and other stakeholders high and dry.41 Businesses destroyed by private equity often operate in sectors like hospitals and newspapers where the damage to communities extends far beyond workers and suppliers.

Private equity’s fee structure incentivizes risk because general partners reap a share of gains when gambles pay off but are largely insulated from losses, which are borne by lenders and other investors. This is exacerbated by the preferential tax treatment of general partners’ share of earnings. Experts have also expressed alarm over the rapid expansion of unregulated private credit, which poses a threat to financial stability.

The proposed rule comes at an especially bad time. Agency understaffing, weakened enforcement, federal legislation, and a Supreme Court decision in Anderson v. Intel could exacerbate the potential effects of the rule.

Experts warn that deregulation could fuel a speculative bubble like the one in the roaring 1920s.42 When the bubble pops, everyone will pay, whether they were playing or not. As University of Chicago Law School Professor William Birdthistle, the former director of the SEC’s Division of Investment Management, warns:

The administration is…encouraging individual retirees to vouchsafe their life savings to exotic financial offerings like private equity. Private equity is, as the name suggests, notoriously opaque, which means retirees would know little about what they’re investing in. The White House and the private fund lobby argue that this policy will “democratize” access to alternative assets and promote “better returns.” But such a plan, which comes with neither the information nor the protections needed to defend investors from serious economic risks, is as compelling as a plan to “democratize” brain surgery. 43

The proposed rule is supported by a financial industry seeking new investors and plan sponsors hoping to reduce litigation risk, not retirement savers.

Whether or not alternative investments have performed well in the past, market saturation, higher interest rates, and other factors will likely reduce future returns. A shrinking client base has made it hard for private equity funds to exit their investments and return funds to clients. Investors in private credit funds have also become skittish due to concerns about lending standards and valuations, prompting some firms to restrict redemptions.44

A survey conducted on behalf of AARP found that “Americans have little interest in adding private market investments and cryptocurrency to workplace retirement accounts.”45 As institutional and wealthy investors try to o􀆯load underperforming funds, retirement savers will be given access to the dregs even as overall quality declines. Retirement savers will be “buying a pig in a poke”—assuming they are even aware they are buying a pig.

We need to better regulate alternative investments, not to loosen regulations protecting retirement savers.

Financial regulations, including disclosure requirements and fiduciary rules, serve multiple purposes. They protect investors, prevent systemic risks such as bank runs, and disclose information needed for financial markets to direct capital to productive uses, rather than activities that do not promote economic growth but simply transfer wealth to insiders from those with less information like most retirement savers and small investors.

Without reliable and comparable information, it is difficult for even sophisticated investors to know whether alternative investments are worth their high cost. We need better regulations to help all investors make informed decisions and guard against conflicts of interest; to fix incentives that encourage value-destroying business practices by private equity and other underregulated financial industries; and to curtail abuse of tax-favored plans by wealthy investors, who have an incentive to load 401(k) accounts up with assets that are difficult to value in order to skirt contribution limits and take maximum advantage of tax subsidies tied to investment returns.

Rather than weakening protections for retirement savers, DOL should work with other agencies to regulate private markets to enable all investors to make informed decisions and protect the economy.

For these reasons, I urge EBSA to withdraw the proposal. Working families need retirement income security, not risky and costly investments ill-suited for small investors.

Thank you for considering my comment.

Respectfully submitted,

Monique Morrisey, Ph.D.
Senior Economist
Economic Policy Institute

1. AARP, “ERISA Litigation Reform: Myth vs. Fact,” accessed June 1, 2026. https://www.aarp.org/content/dam/aarp/politics/advocacy/2026/03/erisa-litigation-reform-myth-v-fact.pdf

2. DOL, “Fiduciary Responsibilities,” web page accessed May 25, 2026. https://www.dol.gov/general/topic/retirement/fiduciaryresp

3. Paul Samuelson, “Risk and Uncertainty: A Fallacy of Large Numbers,” Scientia, Vol. 98, pp. 108-113. https://www.casact.org/sites/default/files/database/forum_94sforum_94sf049.pdf. John Rekenthaler, “How Time Horizon Affects the Odds of Equity Investing,” Morningstar, October 19, 2023. https://www.morningstar.com/columns/rekenthaler-report/how-time-horizon-affects-odds-equity-investing

4. Louis J. Campagna, Information Letter 06-03-2020, EBSA, June 3, 2020. https://www.dol.gov/agencies/ebsa/about-ebsa/our-activities/resource-center/information-letters/06-03-

5. SEC Office of Compliance Inspections and Examinations, Risk Alert, June 23, 2020. https://www.sec.gov/files/Private%20Fund%20Risk%20Alert_0.pdf

6. DOL, “Supplement Statement on Private Equity in Defined Contribution Plan Designated Investment Alternatives,” December 21, 2021. https://www.dol.gov/agencies/ebsa/about-ebsa/our-activities/resourcecenter/information-letters/06-03-2020-supplemental-statement

7. DOL, “US Department of Labor Rescinds 2021 Supplemental Statement on Alternative Assets in 401(k) Plans,” August 12, 2025. https://www.dol.gov/newsroom/releases/ebsa/ebsa20250812

8. SEC, “Staff Statement on Meme Coins,” February 27, 2025. https://www.sec.gov/newsroom/speechesstatements/staff-statement-meme-coins

9. Gary Gensler, “Statement on the Approval of Spot Bitcoin Exchange-Traded Products,” January 10, 2024. https://www.sec.gov/newsroom/speeches-statements/gensler-statement-spot-bitcoin-011023

10. Board of Governors of the Federal Reserve System, “Federal Reserve Board announces the withdrawal of guidance for banks related to their crypto-asset and dollar token activities and related changes to its expectations for these activities,” April 24, 2025. https://www.federalreserve.gov/newsevents/pressreleases/bcreg20250424a.htm

11. EBSA, “401(k) Plan Investments in “Cryptocurrencies,” Compliance Assistance Release No. 2022-01, March 10, 2022. https://www.dol.gov/agencies/ebsa/employers-and-advisers/plan-administration-andcompliance/compliance-assistance-releases/2022-01

12. SEC Division of Corporation Finance, “Staff Statement on Meme Coins,” February 27, 2025. https://www.sec.gov/newsroom/speeches-statements/staff-statement-meme-coins

13. SEC, “SEC Announces Dismissal of Civil Enforcement Action Against Coinbase,” February 27, 2025. https://www.sec.gov/newsroom/press-releases/2025-47

14. Caroline Crenshaw, Remarks at the Investment Company Institute’s 2025 Investment Management Conference, March 26, 2025. https://corpgov.law.harvard.edu/2025/03/26/remarks-by-commissioner-crenshaw-at-the-investmentcompany-institutes-2025-investment-management-conference/

15. Caroline Crenshaw, Remarks at the Investment Company Institute’s 2025 Investment Management Conference, March 26, 2025. https://corpgov.law.harvard.edu/2025/03/26/remarks-by-commissioner-crenshaw-at-the-investmentcompany-institutes-2025-investment-management-conference/

16. SEC, “SEC Enhances the Regulation of Private Fund Advisers,” August 23, 2023. https://www.sec.gov/newsroom/press-releases/2023-155

17. United States Court of Appeals for the Fifth Circuit, “National Association of Private Fund Managers et al. v. Securities and Exchange Commission,” June 5, 2024. https://www.govinfo.gov/content/pkg/USCOURTS-ca5-23-60471/pdf/USCOURTS-ca5-23-60471-0.pdf

18. American Federation of Teachers, Americans for Financial Reform Education Fund, and American Association of University Professors, “From Public Pensions to Private Fortunes: How Working People’s Retirements Line Billionaire Pockets,” July 2025. https://ourfinancialsecurity.org/resources/publicpensionsprivatefortunes; Stephen Deane, “Private Markets: Governance Issues Rise to the Fore,” CFA Institute Research and Policy Center, June 2024. https://rpc.cfainstitute.org/sites/default/files/-/media/documents/survey/private-markets-governanceissues-rise-to-the-fore.pdf; Alexander Ljungqvist, “The Economics of Private Equity: A Critical Review,” CFA Institute Research Foundation Literature Review, 2024. https://rpc.cfainstitute.org/sites/default/files/-/media/documents/article/rf-brief/economics-of-private-equity.pdf.

19. Holger Spamann, Indirect Investor Protection: The Investment Ecosystem and Its Legal Underpinnings, Journal of Legal Analysis, Volume 14, Issue 1, 2022, Pages 17–79, https://doi.org/10.1093/jla/laac003

20. Jeffrey Hooke and Michael Imerman, comment on proposed regulation RIN 1210 AC 20, April 13, 2026. https://www.regulations.gov/comment/EBSA-2026-0166-6564

21. Jeffrey Ptak, “Look Before You Leap When Investing in Private Funds,” Morningstar, July 29, 2025. https://www.morningstar.com/funds/look-before-you-leap-when-investing-private-funds; Linge Sun and Nicholas Reade, “Performance Dispersion in Alternative Asset Classes, CAIS, November 18, 2022. https://www.caisgroup.com/articles/performance-dispersion-in-alternative-asset-classes

22. Stephen Deane and Olivier Fines, “Cryptoassets: Beyond the Hype,” CFA Institute, January 2023. https://rpc.cfainstitute.org/sites/default/files/-/media/documents/article/industry-research/crypto-beyondthe-hype.pdf

23. SEC, “Staff Statement on Meme Coins,” February 27, 2025. https://www.sec.gov/newsroom/speechesstatements/staff-statement-meme-coins

24. Internal Revenue Service, “Retirement topics – Plan assets,” web page accessed May 25, 2026. https://www.irs.gov/retirement-plans/plan-participant-employee/retirement-topics-plan-assets

25. In 2023, the Treasury Department and Internal Revenue Service (IRS) announced that they planned to issue guidance relating to the treatment of nonfungible tokens as collectibles. IRS, “Treatment of certain nonfungible tokens as collectibles,” Notice 2023-27, March 21, 2023. https://www.irs.gov/pub/irs-drop/n-23-27.pdf; David Block and Davide Levine, “IRS Announces Intention to Issue Guidance on NFTs,” Groom Law Group, March 29, 2023. https://www.groom.com/resources/irs-announces-intention-to-issue-guidance-onnfts/

26. Sharon LaFraniere and David Yaffe-Bellany, “How Prediction Markets and Crypto Firms Steamrolled a Watchdog Agency,” The New York Times, May 24, 2026. https://www.nytimes.com/2026/05/24/us/howprediction-markets-and-crypto-firms-steamrolled-a-watchdog-agency.html

27. Ludovic Phalippou, “An Inconvenient Fact: Private Equity Returns & The Billionaire Factory,” University of Oxford, Said Business School, Working Paper, June 10, 202. https://dx.doi.org/10.2139/ssrn.3623820; American Federation of Teachers et al., op. cit. July 2025.

28. Robert S. Harris, Tim Jenkinson, Steven N. Kaplan, and Ruediger Stucke, “Has Persistence Persisted in Private Equity? Evidence from Buyout and Venture Capital Funds,” Journal of Corporate Finance, Volume 81, August 2023. https://doi.org/10.1016/j.jcorpfin.2023.102361.

29. American Federation of Teachers et al., op. cit. 2025; Dan Chung and Brad Neuman, “Debunking Private Equity Prestige,” Alger Insights, April 2024. https://www.alger.com/Pages/Content.aspx?pageLabel=Insights-Commentary-Debunking-Private-Equity-Prestige

30. Jean-Pierre Aubry, “Public Pension Investment Update: Have Alternatives Helped or Hurt?” Center for Retirement Research at Boston College Issue in Brief 22-20, November 22, 2022. https://crr.bc.edu/publicpension-investment-update-have-alternatives-helped-or-hurt/

31. Jean-François L’Her, Rossitsa Stoyanova, Kathryn Shaw, William Scott, and Charissa Lai, “A Bottom-Up Approach to the Risk-Adjusted Performance of the Buyout Fund Market,” Financial Analysts Journal, 72(4), 36–48, December 27, 2018. https://doi.org/10.2469/faj.v72.n4.1https://www.tandfonline.com/doi/abs/10.2469/faj.v72.n4.1

32. Angela M. Antonelli, “Has the Lack of Asset Diversification in DC Retirement Plans Been a Costly Missed Opportunity?” Georgetown University Center for Retirement Initiatives in Conjunction with CEM Benchmarking, June 2023. https://cri.georgetown.edu/wp-content/uploads/2023/06/GeorgetownCRI-CEm- Benchmarking_Lack-of-Asset-Diversification-CRI-paper.pdf

33. SEC, op. cit., August 23, 2023.

34. U.S. Court of Appeals for the Fifth Circuit, op. cit., June 5, 2024.

35. James Van Bramer, “Unpacking the DOL ‘Safe Harbor’ for Alternative Investments,” Plan Advisor, April 14, 2026. https://www.planadviser.com/unpacking-the-dol-safe-harbor-for-alternative-investments/

36. Amy C. Arnott, Christine Benz, David Reyna, and Jack Shannon, “2026 Diversification Landscape,” Morningstar, April 14, 2026. https://www.morningstar.com/content/csassets/v3/assets/blt9415ea4cc4157833/blta0fddf23ec4df239/69ddb33e160be843c15e5ad3/Diversification_Landscape_2026.pdf

37. United States, Executive Office of the President. Executive Order 14330: Democratizing Access to Alternative Assets for 401(k) Investors, August 7, 2025. https://www.whitehouse.gov/presidentialactions/2025/08/democratizing-access-to-alternative-assets-for-401k-investors/

38. Amy C. Arnott, “Sequence of Returns: What It Means and How to Deal,” Morningstar, August 9, 2021. https://www.morningstar.com/retirement/sequence-returns-what-it-means-how-deal

39. EBSA, “FAQs about Retirement Plans and ERISA,” web page accessed May 25, 2026. https://www.dol.gov/agencies/ebsa/about-ebsa/our-activities/resource-center/faqs/retirement-plans-anderisa

40. SEC Office of the Advocate for Small Business Capital Formation, Annual Report 2024. https://www.sec.gov/files/2024-oasb-annual-report-print.pdf

41. Eileen Appelbaum and Rosemary Batt, Private Equity at Work, The Russell Sage Foundation, May 2014. https://www.russellsage.org/publications/book/private-equity-work

42. Andrew Ross Sorkin, “The Rules of Investing Are Being Loosened. Could It Lead to the Next 1929?” The New York Times, October 13, 2025. https://www.nytimes.com/2025/10/13/magazine/investing-private-equitycrypto-crash-1929.html

43. William A. Birdthistle, “Trump Is Pushing Us Toward a Crash. It Could Be 1929 All Over Again,” The New York Times, November 7, 2025. https://www.nytimes.com/2025/11/07/opinion/donald-trump-great-gatsbyroating-20s-sec.html

44. Maureen Farrell, “New Limits on Investors and a Debt Downgrade Add to Private Credit Woes,” The New York Times, March 24, 2026. https://www.nytimes.com/2026/03/24/business/moodys-private-creditdowngrade.html

45. Bryan Miller, “Americans Have Little Interest in Adding Private Market Investments and Cryptocurrency to Workplace Retirement Accounts,” AARP, November 20, 2025. https://www.aarp.org/pri/topics/work-financesretirement/financial-security-retirement/private-market-and-cryptocurrency-investments/

The ban on state regulation of artificial intelligence (AI) contained in both the House and Senate versions of the Republican megabill is overly broad, dangerous to workers, and out of step with public interest. This provision – a ten-year blanket ban on state and local governments’ ability to protect their residents from the harms of AI – is a reckless giveaway to Big Tech that will have far-reaching consequences for economic fairness, worker power, and public trust. 

Both the Senate and House versions of the provision use an extremely broad definition of AI—including automated decision-making systems – tying the hands of state lawmakers from taking any meaningful role in how AI technologies are being rapidly rolled out in many sectors of society. The Senate version ties the moratorium on regulating AI to federal funding for broadband internet infrastructure – a program on which all 50 states and territories rely to make critical progress to improve connectivity.  

The provision is opposed by a broad, bipartisan coalition – including unions, civil rights groups, state attorneys general, members of Congress across the political spectrum, and the public, who understand it as a rash giveaway to big tech. Banning state regulation of AI gives even more power to a handful of billionaires, while reducing the power of working people and communities. 

Congress can still act to remove this harmful provision and ensure that AI can expand in ways that are responsible, innovative, and grounded in public trust—while protecting the rights of workers, consumers, and communities.

Congress has a responsibility to develop and adopt federal standards that ensure new technologies lead to positive economic outcomes – and to ensure that workers have the power to control how AI and related digital tools are used in their workplaces, ideally  through collective bargaining agreements. 

However, in the absence of federal action, it is critical that states be permitted to step in and act – and those lessons can hopefully inform federal policymaking. All 50 states have been working to regulate uses of AI that harm communities and society. This ban would stop all of that progress in its tracks, blocking commonsense AI laws in development or already on the books.  

Here are a few key ways this unprecedented ban on state action to protect workers and consumers could harm workers and erode public trust:

Make it easier for employers to discriminate

The right to equal opportunity at work is already under threat from the Trump administration’s attacks on federal anti-discrimination protections and enforcement.  Unregulated AI systems could speed up and cement discrimination even further. 

Major employers increasingly rely on predictive AI software and algorithmic analysis to choose who they interview, hire, promote, discipline, or dismiss. We know these untested tools for “automated decision making” can fuel discriminatory outcomes, such as a preference for resumes with white- and male-associated names. 

With no federal guardrails in place, and with federal enforcement on anti-discrimination weakened, banning state action would allow discrimination to flourish unchecked. States must be able to step in to address algorithmic bias and enforce anti-discrimination protections so that everyone has a fair shot at a good job. 

Make it easier for employers to drive down wages

With a low federal minimum wage, rampant misclassification of contract workers, and no guardrails on how employers use AI and algorithms to make decisions about pay, the race to the bottom already experienced by gig workers could become the norm in all industries. 

Employers already use algorithms and automatic decision systems to dynamically determine the lowest possible pay for each task, location and individual, with little transparency for workers. If states are blocked from even investigating wage suppression by algorithm, these exploitative practices will spread across all industries.

Increase retaliation, union-busting, and surveillance

Automated surveillance systems and AI-powered monitoring can track everything from workers’ keystrokes and voices to their precise location in their workplace. These tools can be weaponized against workers who organize, speak up about unsafe conditions, or simply take too long in the bathroom. Workers already are vulnerable to unfair – and sometimes illegal – retaliatory discipline and firing. If workers don’t even know the extent to which they are being surveilled, they will struggle to exercise their legal rights or defend themselves from wrongful termination – especially the majority of workers who lack the protections of a collective bargaining agreement.

As one school bus driver told researchers from UC Berkeley Labor Center:

“The bus cameras are the worst— they were originally installed to protect the kids, but now three cameras are pointed directly at us and recording at all times, even when no kids are on the bus. We know now that they use this footage in personnel matters, they listen to us through the bus cameras, and that they use the cameras to read our text messages when we are parked and using our phones while the children are off the bus and we are on breaks from work.”

Preventing states from regulating this kind of surveillance will leave workers more vulnerable to unlawful retaliation and to employer wrongdoing, unsafe conditions, and unfair wages.

Worsen worker privacy

A decade of unregulated AI will allow the aggregation and sale of vast amounts of highly specific data on individual workers in ways that can never be truly erased. For instance: With the aid of AI, data collected from GPS systems and wearable technology could be used to identify an employee’s private medical conditions, even before the worker has the chance to invoke the protections of the ADA or FMLA. If this data is sold to other hiring managers or the open market without any regulations, that same individual will find it difficult to secure future employment. 

Workers and consumers need transparency and tools to  control how AI-powered systems use their personal data – not a decade-long ban on state oversight.

Steal creative work

Artists, writers, musicians, and performers are already seeing their work scraped and reused by AI systems without consent or compensation.  The ban would make it more difficult for creative workers to protect their work, including their own images and voices. Those whose work is stolen without compensation to fuel large language models may have no recourse, even as big tech companies continue to profit. 

Harm public safety and public services

In critical sectors like healthcare, education, and transportation, AI systems are already being used to override expert human judgment. The ban would restrict the ability of workers and their advocates to respond. For example, in healthcare, nurses are fighting to protect patients and provide the care they know is best – even when an algorithm advises otherwise. This is equally true in education, childcare, public safety and transportation – all fields with vulnerable lives and worker safety at risk. 

A better alternative is possible

Corporations do not need a blank check and a deregulated landscape to succeed in creating and selling artificial intelligence, automated decision systems, and related technologies.  The balance of power already tilts too far in favor of employers. Congress should remove this dangerous 10-year preemption of state action from the budget megabill, which already poses serious harm to low-income people in this country. Instead, policymakers should consider responsible AI policy frameworks  through the normal legislative process, where these critical issues can be debated and assessed fairly. 

The Trump administration is taking a reckless approach to deregulation. In his first day in office, Trump ordered a regulatory freeze, barring departments and agencies from issuing any new regulations and pushing back pending regulations until they could be reviewed by a Trump administration appointee. The administration has issued several additional deregulatory orders, including rescission of Biden administration actions that sought to modernize the regulatory process.

These deregulatory executive orders include, among several others:

• EO 14148: Initial Rescissions of Harmful Executive Orders and Actions
• EO 14192: Unleashing Prosperity Through Deregulation
• EO 14219: Ensuring Lawful Governance and Implementing the President’s “Department of Government Efficiency” Deregulatory Initiative
• EO 14294: Fighting Overcriminalization in Federal Regulations

EO 14129’s directive to cut 10 regulations for every new regulation added signals a disdain for regulation as a concept, rather than an appreciation for government efficiency.

Trump is also changing how several institutions responsible for developing and enforcing regulations function, either by reinterpreting their purpose to be deregulatory (as in the cases of the Occupational Safety and Health Administration [OSHA], Food and Drug Administration [FDA], and Environmental Protection Agency [EPA]), or hobbling the effectiveness of entities like the National Labor Relations Board, Equal Employment Opportunity Commission, and Consumer Financial Protection Bureau). This uncritical hostility toward regulation removes guidelines and guardrails that keep U.S. workers and their families safe. The economy is more vulnerable to downturns and public health crises without the protections provided by health and safety regulations.

Why is this happening?

The Trump administration has made its economic policy priorities clear: cut taxes for the rich, make it cheaper to do business, and reduce the scope of government and corporate accountability to workers and their families, especially to poor communities and communities of color. Health and safety regulations prevent employers from engaging in cost-cutting behavior at the expense of workers and communities. Reckless deregulation serves the purpose of reducing costs for employers—even when those regulations protect workers and their families—because regulatory compliance can be costly.

Regulation ensures that workers and their families are shielded from the harms stemming from unrestricted market activity. Policymakers regulate in the wake of crises and tragedies to protect future generations and ensure history does not repeat itself. The Trump administration deregulates recklessly because it dismisses history and does not care about the consequences of bad policy for queer communities, communities of color, or the non-rich, so long as the cost of doing business falls. Inadequate regulation creates the conditions for abuse—and when vulnerability and opportunity for abuse meet, we inevitably see harm. 

Why does that matter for public health and worker safety?

Responsible regulations impose costs on harmful business practices, without which there would be little incentive for employers to prioritize worker and public safety over profit. The FDA requiring milk to meet quality standards before it can be sold in stores imposes a cost on dairy farmers, who must operate their farms adhering to safety standards that result in a higher quality product. Reducing FDA staff such that milk quality testing is no longer feasible reduces the scope for regulatory compliance and corporate accountability to consumers, and raises the possibility of exposure to pathogens including bird flu.

OSHA and EPA regulations ensure that employers will be held accountable when they expose workers and their communities to harmful substances. These regulations require coal mines to protect workers from exposure to black lung disease by providing protective equipment and ensuring their operations are minimally damaging to the surrounding community’s water sources through wastewater management. Removing these health and safety regulations and dismantling the institutions that enforce them certainly makes it cheaper to operate the coal mine, but only through imposing a larger public health cost on miners and their communities. Reckless deregulation may save corporations money and “cut government spending” but delivers the opposite of efficiency.

Why does that matter for racial health disparities?

Black and brown workers and their communities will face the worst consequences of Trump’s reckless deregulatory crusade. Their historical lack of access to wealth and political power means that, without the protections that health and safety regulations provide, they remain vulnerable to irresponsible business practices and policy.1 Laws and regulations prohibiting discrimination in health care and allowing lawsuits based on disparate impact are critical in the battle to close racial health gaps; the Trump administration is dismantling this legal framework through executive orders, setting back that effort toward equity.

Occupational segregation means that minoritized workers are also more likely to be employed doing dangerous work with higher fatality rates and for which health and safety regulations can save lives. In 2023, Black and Hispanic workers died on the job at rates of 3.6 and 4.4 per 100,000 workers, respectively—higher than the rate for workers overall (3.5 per 100,000 workers), and considerably higher than the rates for White and Asian workers (3.2 and 1.6 per 100,000 workers).

Black and Hispanic workers are more likely to die due to heat-related illness on the job compared with white workers, for example. The Trump administration’s regulatory freeze puts a pause on OSHA’s proposed rulemaking on “Heat Injury and Illness Prevention in Outdoor and Indoor Work Settings,” meaning that there are still no federal heat standards for workers. This rule would have required OSHA-compliant employers to develop and evaluate plans to protect workers from heat-related injuries and illnesses.

Losing environmental protections means exposing Black and brown communities to pollution, and further removing means to hold corporations accountable for the pollution they cause. Black and brown communities are disproportionately exposed to toxic waste; when given the funds to properly collect data and enforce policy, institutions like EPA provide the necessary regulatory framework for moving toward environmental justice and ultimately addressing the health disparities that arise from environmental racism. In the face of clearly visible racial inequities, the lack of rules and regulations to address those inequities essentially maintains structural racism.

What will it mean economically for workers and their families?

Reckless deregulation benefits employers and the wealthy at the cost of exposing the rest of the economy to greater risk and inefficiency. Short-term profits and cost savings from deregulation have historically set the stage for devastating economic consequences for workers and their families. Profit-seeking deregulation in banking set the stage for the 2008 financial crisis that tanked the world economy. Predatory payday lenders, exorbitant credit card late fees, and discrimination in lending all represent significant costs to U.S. households.

The greater health risks that come with removing health and safety regulations will cost workers and their families economically. While the Trump administration is removing COVID-19 vaccine mandates from schools, new, more contagious strains of the disease are appearing across the United States. The COVID-19 pandemic is estimated to have cost the United States $14 trillion due to reduced economic output and worsened health and premature death. Ironically, one area where the Trump FDA is pursuing more stringent regulation is in vaccine access. Using the regulatory system to pursue ideological crusades is dangerous and inefficient.

What can we do about it?

The Trump administration is engaged in a reckless crusade against health and safety regulations that will worsen our public health, cost us economically, and ultimately make the country more vulnerable to the next economic downturn and public health crisis. This is happening because the administration is more concerned with providing tax breaks for the wealthy and allowing corporations to irresponsibly cut costs than with protecting U.S. workers and their families. The consequences of this deregulatory crusade will fall heaviest on Black, brown, and poor workers and communities, because they are most in need of the protections that health and safety regulations provide. But make no mistake: The entire economy will suffer in the event of a recession or pandemic brought on by our collective exposure to more risk.

We need a responsible approach to regulation that takes seriously the need for guidelines and guardrails around doing business and implementing policy. A responsible regulatory framework would center equity-enhancing regulations and support the institutions charged with enforcing those regulations. The Trump administration is targeting and dismantling those very institutions. Protecting workers and their families means repairing our damaged regulatory system, being protective of existing responsible health and safety regulations, and being proactive about establishing new regulations where necessary to safeguard against future crises. Trump’s irrational hostility towards regulation benefits the few at the cost of putting us all at greater risk.

1. The subprime mortgage crisis is clear example of this; in the lead-up to the 2008 financial crisis, Black families were targeted with subprime mortgage loans even when they could afford conventional ones because 1) Black families did not have access to mortgages at the same rate as white families and so were eager to own a home when offered a chance (vulnerability), and 2) there were weak regulations on the kinds of mortgages that could be offered to prospective homebuyers and that could be bought and sold in financial markets (conditions for abuse).

Charles Ezell
Acting Director, U.S. Office of Personnel Management
1900 E Street, N.W.
Washington, DC 20415

Re: Improving Performance, Accountability and Responsiveness in the Civil Service (OPM-2025-0004 / RIN 3206-AO80)

Dear Acting Director Ezell:

We write to submit this comment on behalf of the Economic Policy Institute. The Economic Policy Institute (EPI) is a nonprofit, nonpartisan think tank working for the last 30 years to counter rising inequality. We study the impacts of economic policy decisions on low- and middle-income working families at the federal, state, and local level, and our work has also long centered the importance of a strong public sector and well-functioning federal agencies to protecting economic security for the working class. EPI submits this comment to express our strong opposition to the rule as proposed.

The proposed rule, as directed by the reinstated Executive Order 13957,1 outlines a process for revoking civil service protections for certain federal workers who traditionally discharge their responsibilities consistent with the priorities of the administration in office. This directs that civil servants in “positions of a confidential, policy-determining, policy-making, or policy-advocating character” should be classified as “Schedule Policy/Career” (previously known as “Schedule F” in the first Trump administration). EO 13957 outlines steps for agencies to follow in recommending positions to reclassify to the OPM director. The director, in turn, must recommend to the president positions to be reclassified to Schedule Policy/Career in accordance with OPM’s corresponding guidance memorandum,2 with the president making the final decision regarding positions to classify as Schedule Policy/Career. In the following comment, we provide further detail on our highest priority concerns with this rule as proposed.

The proposed rule risks inappropriate politicization of the federal workforce and suggests no meaningful guardrails against this.

The proposed rule claims that employees in this category will not have to personally or politically support the president. However, there are no serious guardrails proposed in the rule to prevent this from happening. If federal employees could be fired at will—without any due process—it would not be difficult for this administration, or any future presidential administration, to target and terminate employees based on political views while claiming whatever justification they choose for doing so. Indeed, now-Vice President Vance once said in a 2021 interview that he would recommend that President Trump “Fire every single midlevel bureaucrat, every civil servant in the administrative state, replace them with our people.”3 Any duly elected President certainly has the right to place a limited number of people into positions of authority across the federal government to help enact their policy agenda. But we have broader civil service protections precisely to prevent this from happening too widely – and before these protections were put into place, government employees were able to be assessed not based on merit, expertise in their jobs, or performance of their duties, but on adherence to a particular political ideology, partisan support, loyalty to individuals, nepotism, or favoritism. Politicizing the federal workforce, or forcing federal employees to toe a certain political line to avoid termination, will further undermine the trust of the American people in the public goods and services that their tax dollars support.

A December 2023 review of research4 on the impacts of greater politicization of public services highlights the significant adverse consequences likely to result from removing civil service protections from policy-influencing employees, including:

• Greater federal workforce turnover, with changes in presidential administrations ushering in new cadres of policy-influencing employees, causing increased instability in agencies’ interpretations and implementations of laws;
• Decline in the institutional history and technical expertise career civil servants develop and bring to bear in formulating policy;
• Less accountability to Congress regarding the laws that agencies administer; and
• More opportunities for political favoritism in federal contracting.

The proposed rule could present serious risks to accurate data collection and analysis.

The proposed Schedule Policy/Career designation also threatens the integrity of federal data and the federal statistical system. This is particularly concerning to EPI given our role as a research institution: Conducting credible economic research requires access to reliable public data sources. The federal government has a well-deserved reputation for gathering and publishing high-quality economic data, including open access data that policymakers and policy-oriented organizations like ours rely on for economic decisionmaking and research. Statistical agencies like the Bureau of Labor Statistics and Census Bureau require highly-trained staff to produce impartial and reliable data, but the Schedule Policy/Career designation would eliminate the norms and incentives critical to producing accurate data. Without professional autonomy, the statisticians and economists who staff these agencies will leave, employee productivity and morale will decline, and the remaining staff will face pressure to alter or suppress data in accordance with a political agenda. Career staff are also less likely to risk their jobs to leak embargoed market-moving data to insiders who can use the information to their advantage at the expense of other investors.

Businesses also require high-quality data to make sound investment decisions. A 2018 National Association of Business Economics survey of private sector members highlighted the vast importance of government data for the private sector, according to an analysis by Hughes-Cromwick and Coronado (2019).5 When asked “Are government data important to analyses and forecasting that drive business decisions,” 95 percent of respondents replied “yes.” At least half of respondents said that government data inputs were “critical” to each of the following activities: capital spending decisions, pricing decisions, financing decisions; and interest rate and asset allocation decisions. Over 90 percent of respondents used government data to inform business decisions from the Bureau of Economic Analysis, Bureau of Labor Statistics, Federal Reserve, and Census Bureau.

There is no substitute for data from the federal government. Even many private-sector data alternatives ultimately rely on the official data produced by the federal government. For example, the ADP National Employment Report is benchmarked to the Quarterly Census of Employment and Wages.6 The politicization of federal statistical staff will undermine public trust in official data.

The proposed rule presents risks to government services and to the continuity and stability of the federal workforce.

Acting Social Security Administrator Leland Dudek recently outlined his plan to reclassify the staff of entire offices of the Social Security Administration under the Schedule Policy/Career designation in this proposed rule.7 Acting Administrator Dudek’s plan, in our view, would warp any reasonable interpretation of the term “policy-influencing,” as it would include staff who research and publish statistics of the agency’s performance, work in human resources, maintain the agency’s information technology and database operations, and adjudicate disability claims. As the Center on Budget and Policy Priorities has noted, the SSA is one of the most nonpartisan and apolitical federal agencies, “with one of the highest ratios of civil servants to political appointees” in federal government – in 2024, just 19 political appointees out of 57,000 total employees.8 Radically changing this ratio to expand the share of political appointees would risk politicizing one of our most critical federal services, and exposing that many staff to the risk of at-will termination would exacerbate the long wait times, website crashes, and other operational disruptions that have already characterized the last few months at the agency. It could also lead to pressure on adjudicators to limit access to benefits. This would cause further harm to the millions of senior citizens, people with disabilities, and low-income individuals in the U.S. who rely on Social Security benefits and need reliable, trustworthy communication with SSA staff and infrastructure.9

We are concerned that this rule could be used not to improve accountability or responsiveness in the federal workforce, but to further carry out initiatives started by the White House’s new office of the Department of Government Efficiency (DOGE). DOGE and its leadership have undertaken actions to reduce the size of the federal workforce and find related cost savings by any means necessary, with little regard for actually assessing the significance or quality of the work performed or the critical services being provided by those federal workers. This has included attempts to shutter whole federal agencies authorized by Congress, place employees on administrative leave, offer deferred resignation to employees, and to disregard the terms of collective bargaining agreements with federal employees. There is ongoing litigation against these actions from the executive branch – for example, recently a federal judge issued a restraining order in American Federation of Government Employees, AFL-CIO, et al. v. Donald J. Trump, et. al.10 blocking the administration from making further reductions in force while the litigation is in progress to determine the legality of the administration’s actions. Advancing this rule could provide more cover for reclassifying many positions to be more easily terminated, bypassing other checks and balances on this authority and having a chilling effect on whistleblowers.

We oppose this rule as proposed and urge the agency to withdraw the proposed rule. We also urge this administration to respect existing laws – and collective bargaining agreements, where relevant – with regards to hiring, firing, and changing employment conditions for career civil servants and excepted service employees alike.

Sincerely,

Samantha Sanders
Director of Government Affairs & Advocacy
Economic Policy Institute

Ben Zipperer, Ph.D.
Senior Economist
Economic Policy Institute

1. White House. 2025. “Restoring Accountability To Policy-Influencing Positions Within the Federal Workforce” (Executive Order). January 20, 2025.

2. U.S. Office of Personnel Management (OPM). 2025. Memorandum, Subject: Guidance on Implementing President Trump’s Executive Order titled, “Restoring Accountability To Policy-Influencing Positions Within the Federal Workforce.“ January 27, 2025.  

3. Prokop, Andrew. 2024. “J.D. Vance’s radical plan to build a government of Trump loyalists.” Vox, July 18, 2024.

4. Moynihan, Don. 2023. “The risks of Schedule F for administrative capacity and government accountability.” (Commentary), Brookings Institute website. December 12, 2023.

5. Hughes-Cromwick, Ellen and Julia Coronado. 2019. “The Value of US Government Data to US Business Decisions,” Journal of Economic Perspectives, 33(1): 131-146.

6. ADP. 2025. Technical Notes to the ADP Employment Report, retrieved May 15, 2025.

7. Wagner, Erich. 2025. “Dudek calls for entire SSA offices to be converted to new Schedule F.” Government Executive, April 22, 2025.

8. O’Connor, Devin and Romig, Kathleen. 2025. “Trump Administration Politicizing Social Security Administration Will Further Undermine Benefits.” Off the Charts Blog (Center on Budget & Policy Priorities), May 8, 2025.

9. Perez, Daniel and Morrissey, Monique. 2025. “Southern and Midwestern districts are the most vulnerable to Social Security cuts and disruptions.” Working Economics Blog (Economic Policy Institute), May 13, 2025.

10. Order Granting Temporary Restraining Order and Compelling Certain Discovery Production, American Federation of Government Employees, AFL-CIO, et al., v. Donald J. Trump, et. al. (N.D. Cal 2025) (No. 3:25-cv-03698-SI).

This report provides a brief overview of what is known so far about the economics of AI, highlights some accounts of how it is being deployed in anti-worker ways across different industries, offers a landscape analysis of federal legislation, and then presents policy recommendations on what federal legislation should be aiming to achieve given the latest research on the likely impacts of AI on workers and our economy. A key theme of the policy recommendations is to keep sight of the broader economic and institutional contexts in which AI might be deployed and to avoid tunnel vision in crafting policies that are too narrowly focused on the latest tools used by employers (rather than the underlying harmful practice).

The deployment of AI-powered technologies in the workplace has been associated with an array of worker experiences and outcomes that warrant concern, including discrimination, unsafe working conditions, seemingly arbitrary disciplinary action or discharge, among others. This has led many policymakers and advocates to put AI on the top of their priority list for drafting new legislation or regulations meant to protect workers. Policymakers have generally focused on assessing how the introduction of AI might degrade the on-the-ground experience of workers. While these assessments are illuminating, they have often steered policy responses toward solutions that are tailored to current forms of AI technologies or use cases. This approach, however, leaves policymakers in the position of potentially failing to keep up with other tools and practices available to employers that result in the very same outcomes (especially as technologies continue to advance) and runs the risk of falling short of providing meaningful long-term benefits to workers.

For years, employers have used algorithmic or automated systems in ways that harm workers—including through discrimination, diminishing workplace safety and privacy, and limiting decision-making power. And for decades, employers have been empowered to use garden-variety management practices that do all these things. Policymakers should rightly be alarmed by the experience of workers today, but they should look beyond the latest tool businesses are using to achieve these outcomes and instead look deeper to the underlying conditions that enable these results.

At the most simple level, the impact of AI technologies on workers is a function of the balance of power between businesses and their workers. If there is balance in the labor market because policy has empowered workers, then most new technologies (including AI) will be steered into generating productivity gains that can be shared across the economy. If the labor market remains unbalanced and workers disempowered, then lots of new technological tools will instead be deployed in zero-sum ways to keep wages suppressed and the incomes of capital owners and managers high.

The rest of this report builds to a set of recommendations for how policymakers should approach empowering workers in the age of AI. First, we review the likely economic impacts of AI technologies. Second, we provide a landscape analysis of existing federal legislative proposals. Finally, we outline three key pillars of a worker-centered policy strategy in the age of AI that aim to achieve a simple objective: increase the ability of workers to meaningfully engage their employers in how AI technologies are deployed in the workplace. These pillars are: 1) expand and expedite pathways to collective bargaining, 2) reduce AI-specific barriers to worker voice and strengthen employment protections, and 3) increase workers’ ability to leave employers who create exploitative conditions with AI (or other technologies and practices) and ease job transitions.

Labor market impacts of advancements in AI

There has been a great deal of speculation and attention given to the potential labor market impacts of new AI technologies over the last few years. In particular, many people have voiced worries about mass job loss and the impact of technologies on skills required for various occupations.1 Others have raised concerns about AI undermining core workplace rights and altering conditions of employment.2 In this section we identify what we believe are the known labor market impacts of AI. These evidence-based outcomes should serve as guideposts for evaluating current legislative proposals and inform additional policy recommendations.

Aggregate labor market impacts: Employment effects and employer demand for skills

While anecdotes and limited data analysis have fueled conversations about mass job loss, it’s important to consider a longer-run assessment of the impacts of technology on the labor market, drawing on a rich body of economic research and measurable outcomes during prior periods of technological advancement. 

Earlier this year, Bivens and Zipperer (2024) published a thorough assessment of AI’s likely impact on the labor market and workers, given the evidence from past waves of major technological progress. Their analysis highlights that technological advancements and subsequent productivity gains are highly unlikely to lead to mass joblessness.

Concerns that technological advancements will create significant job loss rest on the basic theory that AI technologies will increase productivity such that firms will use less labor, resulting in mass layoffs and a rise in unemployment. However, this theory muddles firm-level experiences with aggregate outcomes and glosses over a crucial piece of the equation: whether there will be other jobs in the labor market available—including those created by technological change—for workers who may be laid off from an individual firm.

Unemployment increases when the potential output of the economy (how much could be produced if nearly all of the labor force were employed) exceeds aggregate demand (total spending by households, businesses, and government). Thus, even if AI-based technologies result in an increase in productivity, whether this will result in an increase in unemployment depends on if aggregate demand doesn’t similarly increase. Bivens and Zipperer (2024) further note that fiscal and monetary policies enable aggregate demand to be boosted much more quickly than potential output can move, allowing policymakers to minimize the magnitude and duration of any disconnect between potential output and aggregate demand.

Relatedly, a review of the relationship between productivity growth and unemployment highlights that productivity gains do not occur at the expense of rising unemployment. In fact, Bivens and Zipperer (2024) find that “fast productivity growth is associated with lower average rates of unemployment across business cycles.” Moreover, their analysis shows that faster productivity growth does not hamper the pace of falling unemployment over a business cycle. In short, there is no compelling evidence to support the theory that AI-induced productivity gains will result in mass workforce displacement. 

Beyond aggregate employment effects, there has been a corollary concern among some policymakers about how businesses’ use of AI technologies may impact the demand for workers without a college degree. That is, as more complex technologies are developed and deployed in the workplace, will firms hire more highly skilled individuals for jobs that previously required lower skills? Here too, the theory for why this may occur is straightforward, but a dig into the data reveals why this is unlikely to occur in practice.

First the economic theory: An increase in use or advancements in technology will lead to an increase in demand for more highly skilled workers relative to workers who don’t have the skill set to use these technologies. Rising demand for higher-skilled workers will, as the theory goes, increase wages among those relative to lower-skilled workers, increasing income inequality. In general, economists have often used a college degree as a proxy for workers who obtain the skills necessary for new technologies. 

As Bivens and Zipperer (2024) discuss in greater length, a review of the data shows that this theory hasn’t played out in practice over the last 20 plus years. Since 2000, there has been little to no change in the college wage premium despite significant technological advancements. In fact, over the last few years the college-to-high-school wage premium has fallen, resulting in a decline in wage inequality between higher- and lower-skilled workers. Therefore, there is little evidence to suggest that continued advancements in AI will trigger a greater need for “upskilling” to counterbalance decreasing demand for lower-skilled workers (Bivens and Zipperer 2024). 

In the period since 1979, there has been a consistent shift in labor market power away from typical workers and toward corporate managers and business owners, resulting in a redistribution of income. However, as Bivens and Zipperer’s analysis shows, a close review of economic research does not support the notion that technological advancements over this period have been the causal driver of worsened outcomes for workers. While technological change may have coincided with rising income inequality, it was itself not a headwind for workers or our economy. Conversely, under the right conditions—more equal distribution of market power between businesses and workers—advances in technology can and have resulted in improved outcomes for workers.

It’s not difficult to see how improvements in technology will help spur greater economic growth by making production processes more efficient. The question then becomes whether workers will be able to share in the gains that flow from new advancements in technology. The answer hinges on the relative labor market power of workers. Given today’s significant imbalances and the erosion of core labor rights, it is unlikely that workers will share fully in the economic gains brought about by AI technologies or any other driver of productivity growth. 

Industry-, firm-, and worker-level impacts of AI 

Beyond aggregate-level impacts, many studies and reporting chronicled the experience of workers using AI and how the use of these technologies have played out within specific industries or occupations. Below we discuss the types of AI technologies that are being used across three industries—warehousing, call centers, and health care; how the specific manner in which they have been implemented has exacerbated negative outcomes for workers; and the efforts workers and their representatives have taken to mitigate these effects.

AI in warehousing

The use of AI technologies in the warehousing industry is perhaps among the most publicly recognized. Over the last few years, there has been extensive documentation of the use of technologies in this industry to set productivity targets and task management, particularly among large retailers like Amazon and Walmart.3 For example, handheld devices track individuals’ pace of work including the number of packages they may be scanning in an hour, error rates, and time between scans (Bernhardt, Suleiman, and Kresge 2021). These data are then used by companies to establish productivity quotas or inform disciplinary decisions. Additionally, robot carts can be used to direct where workers should go in a distribution center, identify which products to move, and set the pace of work (Bernhardt, Suleiman, and Kresge 2021). 

While this is an industry with historically higher injury and illness rates than national averages, these incidents have ticked up in recent years.4 To be sure, not all injuries are a result of using AI technologies, but it is clear that the manner in which these technologies are currently being used in many establishments is associated with severe injuries.

The National Employment Law Project (NELP) found that 1 in 15 Amazon employees experiences a recordable injury each year in an analysis of Occupational Safety and Health Administration (OSHA) injury reporting records for establishments with more than 1,000 workers (Tung, Marquez, and Sonn 2024). Amazon’s injury rate has consistently been higher than the national average and even exceeds other industries with notable workplace hazards such as coal mining, forestry, and logging (Athena Coalition et al. 2020). As the researchers note in an earlier study of OSHA’s records, the most common injuries are to “workers’ backs, shoulders, knees, wrists, ankles and elbows. These types of injuries are often caused by workers assigned tasks involving ergonomic hazards including forceful exertions, repetitive motions, twisting, bending, and awkward postures” (Athena Coalition et al. 2020).

Almost all cases (95%) result in employees missing work or being assigned to different job duties; these injuries can have lifelong impacts and associated problems (Tung, Marquez, and Sonn 2024). The report goes on to explain that “the high rates of serious injury at Amazon are directly attributable to the way that the company manages its workforce using intensive surveillance, automated discipline, and constantly changing quotas generated by algorithms” (Tung, Marquez, and Sonn 2024).

Injuries in the warehousing sector are not unique to Amazon, though research by NELP indicates rates of injuries are far higher at their facilities. Recognizing the prevalence and severity of injuries in the warehousing sector caused by a range of practices, the Occupational Safety and Health Administration launched a national emphasis program (NEP) in 2023 to address comprehensive hazards in this sector (OSHA 2023a). Notably, the NEP’s inspection procedures included an ergonomics screening. Specifically, during interviews with workers, while reviewing injury logs, and site walkthroughs, OSHA inspectors will assess whether workers are exposed to ergonomic hazards and if so, then expanding the scope of investigation (OSHA 2023b). Importantly, in the absence of a specific ergonomics standard, OSHA can only issue citations related to ergonomic hazards against employers under the general duties clause of the OSH Act (OSHA 2024).

In response to poor working conditions, workers and advocates have sought solutions through labor organizing efforts and state-level legislation. While strides are being made in certain states like California, Minnesota, New York, and Washington, new organizing efforts at individual establishments have been arduous. For example, reporting has shown that Amazon may be using electronic monitoring and productivity quotas as a means to retaliate against employees engaged in union organizing. Specifically, in one case at a Kentucky fulfillment center, an employee filed a complaint with the NLRB claiming that Amazon used his failure to meet specific performance targets as pretext to retaliate against him for leading a union organizing campaign (Rosenberg 2022). While Amazon employees at a Staten Island fulfillment center successfully voted to join a union in 2022, they have not yet secured a first contract (Chapman and Hadero 2024).

In an analysis and discussion of injury rates at Amazon, Julia Lang Gordon (2021) explains how workers’ diminished market power—particularly, limited alternative employment opportunities and the prospect of being fired without cause—has enabled “Amazon to push its employees to the physical brink while facing little to no repercussions.” At establishments where workers are able to secure representation by a union and a collective bargaining agreement, there is a clear pathway to improving working conditions, including those that are undermined by AI technologies. It would be exceedingly unlikely, for example, for a unionized employee to be fired for failing to meet unrealistic productivity quotas that if met, would require engaging in unsafe practices; workers would similarly have a clear process to protect basic rights like accessing the bathroom and taking rest breaks.

The warehousing industry is a clear example of how workers’ lack of bargaining power has directly limited their ability to engage businesses in how technologies are used in the workplace and prevent unsafe practices that have resulted from their use. 

AI in call centers 

The impact of electronic monitoring and management on working conditions is not unique to the warehousing sector. In call centers, AI technologies have been deployed to replace, assist, and manage workers’ activities. For example, automated systems are being used to screen calls and route them to employees; chatbots are being used to provide workers with prompts and answers to callers’ questions; other systems are generating notes and transcripts of employees’ calls with customers, identifying patterns in behavior or deviations from call scripts; and other technologies are providing real-time management to employees, including prompts related to their tone, pace of speech, and cues to signal specific emotions (Bernhardt, Suleiman, and Kresge 2021; Doellgast et al. 2023).

In a recent survey of call center workers, researchers explored how the use of these technologies impacted workers’ outcomes. While the experience with these technologies varied, most workers held negative views of AI management tools. About two-third of workers reported that automated monitoring systems made their jobs more stressful and did not feel that monitoring or coaching systems increased fairness on the job (Doellgast et al. 2023).

Among the most concerning impacts is the potential for real-time recording systems to generate performance evaluations, and then inform disciplinary decisions and firing without employees having an opportunity to understand what is driving these decisions, the accuracy of the data underlying them, or avenues for appeal. In a written statement for a U.S. Senate forum on AI, Ameenah Salaam (2023) of the Communications Workers of America describes how CWA workers are experiencing this very dynamic: 

These systems are often ineffective and have negative impacts on the workplace and the quality of service our members can provide. We’ve heard from workers of color about discriminatory bias from systems that purport to judge expressions of emotion, like empathy; workers also say the systems do not recognize certain pronunciations and styles of speech. Agents report that the scripts enforced by these systems slow down the work of helping customers and often advise wrong solutions that may violate company policies, creating a situation where agents can face discipline for following the system’s prompts.  

As with any workplace, the presence of a union enables workers to engage their employers and address how specific business practices affect job quality and their well-being; this is particularly true for call centers where employees have long grappled with electronic surveillance and monitoring. Research on the impact of unions in call centers has found that unions not only increase worker well-being but are able to successfully negotiate over how information gathered from performance monitoring is used (O’Brady and Doellgast 2021). Sean O’Brady and Virginia Doellgast (2021) write, “unions place a strong emphasis on challenging ‘discipline-based’ performance management practices and encouraging more developmental ones focused on training and development, as well as in establishing more fair and transparent processes for evaluating and rewarding performance.” In fact, in their study on the impact of unions on key worker outcomes, the authors found that union activities were positively associated with developmental performance monitoring and a greater perception among workers of the fairness of performance metrics. Overall, the authors found that union engagement improved worker well-being (measured by emotional exhaustion) (O’Brady and Doellgast 2021). Similarly, as Aurelia Glass (2024) has written, the CWA has over the last three decades secured contracts with major companies, including AT&T and Verizon, that have placed limits on the frequency of monitoring employees calls and the ways in which recordings can be used to inform disciplinary action.

Thus, it is clear that to the extent workers and policymakers are concerned about how AI management systems or performance metrics will be used and the fairness of the underlying AI technologies, unions have a clear track record of being able to address these very issues.

AI in health care

In the health care sector, workers have raised significant concerns over how AI-related technologies have impacted staffing ratios, limited practitioners’ discretion in patient care, and resulted in incomplete or incorrect care plans. In a recent survey of registered nurses, more than 50% said their employers utilized an algorithmic system to analyze patients’ health records to assess patient care needs (NNU 2024). The survey, conducted by the National Nurses Union (NNU), also found that a growing number of shift handoff reports from one nurse to the next was partially or wholly generated by AI technologies. And between 30% and 40% of nurses whose employers use AI systems are not able to override or correct care assessments or outcome predictions associated with discharge recommendations (NNU 2024).

The NNU’s research alarmingly finds that there are notable deficiencies in the accuracy of AI systems in use. For example, nearly 70% of nurses who use algorithmic systems to assess patient acuity reported that their own assessment did not match that of the computer. Similarly, nearly 50% of nurses who receive AI-generated handoff reports said their own assessments differ from the information provided to them (NNU 2024). 

Like warehousing, call centers, and other industries where AI technology is being deployed, the use of AI in the health care setting is exacerbating long-standing issues that health care professionals have bargained over, including staffing ratios, and discretion in patient care. While NNU and others are ringing alarm bells for policymakers and the public on the impact of AI on nurses and patient care, the impact of AI on health care professions goes well beyond nurses. Physicians and others who are less commonly represented by labor unions may face greater challenges in raising awareness of these issues and mitigating harmful outcomes.

The three industries discussed above, in addition to research on the use of AI in other industries and occupations, highlight three general themes that should inform policy evaluation and development. First, AI technologies are often the latest iteration of prior technologies and management strategies. While the use of certain AI technologies is becoming standard practice in certain industries and occupations, there remains significant variation in use across firms. Second, there is a lack of transparency around when technologies are being used, how they function, and how they inform employment-related decisions. Third, and perhaps most importantly, under certain circumstances AI-based technologies are undermining existing labor and employment rights. It is important to note, that these three broad findings and trends do not focus on the inherent impact of AI technologies but rather the impact of businesses using these technologies in the context of today’s labor market—most crucially the current balance of power between workers and businesses. Across a range of sectors, unions are negotiating over employers’ use of AI in the workplace and have secured crucial wins for workers (Glass 2024). Given weakened labor laws, though, it is exceedingly difficult for most workers today to successfully secure representation and bargain over AI and other issues.

Given the aggregate-level labor market impact of AI as well as the variety of ways AI technologies are being used across industries and firms, the objective of federal legislation should be to increase the ability of workers to meaningfully engage their employers in how AI technologies are deployed in the workplace.

Landscape of federal legislation 

There is a bevy of federal legislation, principles, and road maps intended to curb negative impacts of AI on workers. In the appendix below, we provide a complete review of these bills, highlighting the most common policy interventions. After reviewing these proposals, it’s clear that few bills include comprehensive solutions for workers in a period of AI advancement. For example, many bills include new employer disclosure requirements, ensuring workers are aware of when certain practices like workplace surveillance and data collection are occurring. However, given workers’ lack of bargaining power and low unionization rates, it may be challenging for the typical worker to adequately leverage this new information on their own to change their employers’ practices.

Notably, one proposal—the Warehouse Worker Protection Act—takes a novel approach by aiming to address head-on unique barriers to organizing in the warehousing industry. The bill does this in two ways. First, it limits the ability of employers to fire workers for failing to meet productivity quotas that haven’t been previously disclosed to the employee and requires greater transparency of data disclosure to allow employees and their representatives to assess whether employment actions based on performance monitoring are consistent across employees. Second, the bill would amend the National Labor Relations Act and the Occupational Safety and Health Act to strengthen the ability of these agencies to enforce workers’ rights. But even here, the amendments are narrowly tailored.

Most of the proposed bills to date aim to address concerning issues workers are facing—discrimination, invasion of privacy, lack of recourse for hiring, disciplinary or other employment actions, and safety risks. But unless AI-specific proposals are paired with solutions to address workers’ bargaining power, these AI policies will fall short of their very objectives. Congress may soon be caught in a game of whack-a-mole, chasing the latest use case of AI systems as employers adjust practices in response to new legislation and technologies, all while working conditions continue to deteriorate.

Policy discussion and recommendations

In the prior sections, we discussed the likely impact of AI on the labor market as well as the current experience of workers who use or are subject to AI-powered technologies. Taken together, these economic realities have led policymakers to question what policy interventions are needed during a period of AI advancement. However, we believe the success of any AI policy will hinge on its ability to address the underlying conditions that have, for decades, been enabling businesses to use the latest tool available to them—in this case, AI technologies—to undermine worker power and erode workplace conditions and outcomes. Fundamentally, the impact of AI on workers is not solely a function of the design of the tool itself but instead the institutions and policies that shape how it is implemented in firms across America.

Workers should have a voice in how policies are implemented in their workplaces. But decades of weakened bargaining power and increased reliance on employers for basic necessities such as health care and retirement have made it exceedingly difficult—if not impossible—for most workers to bargain over how AI technologies are being used in the workplace, let alone walk away from poor working conditions. These broader dynamics have led to the very outcomes policymakers are rightly concerned about when it comes to AI: diminishing workplace safety, discriminatory hiring, limited decision-making power, lack of data privacy, and disciplinary actions without explanation or appeal.

Therefore, any set of AI policies must first address the systematic erosion of core labor rights and meaningful exit options for workers in exploitative job situations. Conversely, to the extent policymakers aim to construct AI-specific policies, they need to do so strategically. This means assessing whether proposed solutions to common outcomes for workers who use or are subject to AI will 1) meaningfully remove unique impediments to the exercise of worker voice on the job or 2) create new protections that will be realized by workers beyond today’s latest version or use of AI systems. Policymakers may be tempted to craft solutions that are narrowly focused on addressing the consequences of the use of AI technologies today. But this approach runs the very real risk of crafting solutions that quickly become ineffective as employers’ practices change or technologies evolve. Instead, a wider lens approach to creating standards and protections that happen to also stand up against employers’ current use of AI technologies will be more likely to meaningfully improve worker outcomes for years to come.

Indeed, Congress has long had to balance this dynamic—deciding whether to regulate an outcome or the mode by which an outcome is realized—when designing some of the nation’s most fundamental employment protections. For example, when Congress passed the Fair Labor Standards Act, legislators debated whether to explicitly include a list of the common employment structures that they believed businesses would use to evade coverage under the FLSA such as piece work, off-premises work, and commissions, among others. As Kati Griffith (2019) writes in a historical analysis of the FLSA:

Congress eventually rejected this much-discussed list of specified devices of evasion in favor of a much broader and more flexible concept of employment. It provided definitions that could adapt with the times and adapt to new “devices.”…By moving away from a specific list of tools of evasion, Congress gestured that it did not want to slide into an endless game of whack-a-mole, to preempt different business structures and strategies that might emerge to sidestep the FLSA’s coverage. 

While not directly analogous to the choices Congress faces today, the legislative history of the FLSA is nonetheless informative as Congress considers how and to what extent they should regulate AI in the workplace. If AI-related outcomes highlight weakness in existing rights (like health and safety) then the foundational statutes, regulations, and enforcement efforts should be strengthened to ensure the outcomes arrived at by the use of AI are indeed protected as intended. Similarly, to the extent AI-related outcomes have shined a light on new or unlegislated risks to workers, then Congress should address the issue broadly; for example, issues related to data privacy should be addressed in a manner that ensures clear lines of privacy in the workplace are drawn for employees and personal information—no matter how it is obtained—is not used for illegitimate purposes.

Below we discuss three pillars of a worker-centered AI policy strategy that draws on the themes discussed above with the aim of achieving a straightforward objective: increasing the ability of workers to meaningfully engage their employers in how AI technologies are used in the workplace. 

Pillar one: Expand and expedite pathways to collective bargaining

Much of workers’ experiences with using or being subject to some form of AI technology today is mostly a function of low unionization and weak bargaining power, not of the current state of technology. Relatedly, efforts to increase transparency around the use of AI technologies (discussed below) will only be effective at catalyzing change if workers and their representatives are able to act upon that information. Therefore, a key pillar of any worker-centered AI policy strategy is expanding and expediting pathways to collective bargaining. The most powerful way to do this would be for Congress to pass the PRO Act and the Public Service Freedom to Negotiate Act (PSFNA). However, this much needed fix to our labor laws has languished in Congress for far too long. 

While considering policies to support workers and create good jobs in an age of AI, policymakers should also look closely at other ways to restore collective bargaining rights and increase bargaining power of workers, even if it is short of the more comprehensive solutions of the PRO Act and PSFNA, or limited to AI-related issues.5 In many other countries, sectoral bargaining is used to reach industrywide agreements on core conditions of work. In the United States, though, sectoral bargaining has not taken hold, but the Clean Slate for Worker Power project and others have put forward a rich set of policy recommendations on how to advance sectoral bargaining in the U.S. (Clean Slate 2021; CLJE Lab 2024a). For example, tripartite boards composed of representatives from employers, workers, and the public could be convened to negotiate new standards for specific industries or issues (Madland 2019). Congress should consider whether emerging dynamics related to AI may be ripe to address through sectoral bargaining, which is particularly useful in sectors in which there is already strong union representation or, conversely, high levels of monopsony power. 

Pillar two: Reduce AI-specific barriers to worker voice and strengthen employment protections

Workers in the United States—particularly nonunion workers—are far too constricted in how they can express voice over how their workplace is organized and run. One key impediment is the opaque management practices and production processes in workplaces. AI—a highly specialized technology that very few understand deeply—threatens to make expressing voice over workplace practices even harder.

Policymakers have, therefore, focused much of their attention on using legislative interventions to increase firm-level awareness and knowledge of technologies that are in use. However, it’s important to realize that legislative solutions that fill the knowledge gap that results from weak bargaining power is only part of the equation. In other countries with strong labor unions, employees have been able to use their market power to not only extract a clear understanding of what technologies are being used, but to use that information to further engage businesses to bargain over how technologies should be used. It is this latter part of the equation where current federal proposals seem to fall short. If, in the absence of adequate worker bargaining power that would allow workers to effectively demand transparency from employers on their own, Congress is attempting to step in and require disclosure of technological practices, then these proposals should be created in a way that maximizes the ability of labor unions, worker advocates, and nonunion employees to realistically use that information.

Therefore, policies that include new employer disclosure requirements to workers should also create a public repository of businesses’ disclosure reports. Greater public access to this information will reduce barriers for labor unions, worker advocates, and researchers to access the information, enabling these entities to identify and take action on industry trends and worker outcomes. Similarly, proposals that create new data privacy protection for workers, which often include a right for workers to receive and review their data, should also consider how workers could realistically understand and act upon this information. Other experts, for example, have suggested the use of AI monitors in the workplace, which would help create a venue for workers—union and nonunion alike—to better understand disclosures provided by employers and engage with them directly on complex issues (CLJE Lab 2024b).

Beyond increasing transparency related to the use of AI, there are a number of existing labor and employment laws that limit its use. While these existing laws give workers some voice and power in their workplaces, it is crucial to ensure that the laws are not subverted by the deployment of AI in workplaces.

The EEOC, NLRB, and DOL have all issued various forms of guidance clarifying the applicability of relevant laws to the use of AI technologies. The EEOC has multiple resources for employers related to the use of AI in hiring-related decisions and is actively engaged in enforcement actions against businesses that have violated the law (EEOC 2023). Similarly, the NLRB general counsel has issued a memorandum identifying how the use of surveillance and monitoring technologies may violate core employee rights under the NLRA (NLRB 2022). Finally, the Department of Labor’s Wage and Hour Division has issued a Field Assistance Bulletin to their investigators, clarifying how AI technologies may result in violations of minimum and overtime standards, as well as employees’ rights under the FMLA, and other laws (DOL 2024).

Looking ahead, policymakers should first and foremost ensure enforcement agencies are adequately funded. While the size of the workforce and complexity of businesses continue to grow, for too long agencies’ enforcement capacities have not kept up. Enforcement agencies will be limited in their ability to ensure employers’ use of AI technologies does not run afoul of existing laws if Congress does not provide them with adequate funding. Relatedly, Congress should consider what new requirements might aid in the compliance and enforcement actions under existing laws. For example, policymakers should leverage the authority and expertise of existing federal entities, such as EEOC, NIST, or the FTC, to standardize auditing frameworks and tools that can be easily used for pre-deployment and ongoing testing. The first-order objective of these audits should be to identify and eliminate outcomes that violate existing labor and employment laws. Additionally, as Congress considers new proposals, whether they are disclosure requirements or new employment standards, policymakers should take steps to ensure there are avenues for workers and their representatives to more fully participate in the enforcement process, including a private right of action.

Additionally, policymakers and federal agencies should consider how to more fully leverage existing authorities to address workplace outcomes that have been exacerbated by the use of AI technologies by issuing new guidance and regulatory standards. For example, OSHA should consider which industry-specific standards would better protect employees from workplace injuries that can arise when certain AI technologies are deployed in a specific manner; NIOSH should study the mental health impacts of specific working conditions and practices, including those with AI technologies, creating the evidentiary basis of potential OSHA standards. Finally, where AI has shined a spotlight on unregulated harms to workers that warrant broader interventions like issues related to data privacy or invasive workplace surveillance, Congress should advance new employment standards.

Pillar three: Increase workers’ ability to leave employers who create exploitative conditions with AI (or other technologies and practices) and ease job transitions

In earlier sections, we discuss how aggregate job loss is unlikely to occur as a result of advancements in AI technologies. For each job displaced by AI, there is highly likely to be one created by the ripple effects it creates (as has happened with other technological changes).  However, job churn creates stress and income losses for workers in the sector seeing job displacement from AI (Bivens and Zipperer 2024). If Congress is serious about easing the impact of AI-related job transitions for workers, then the most effective solution is strengthening social insurance programs to reduce the economic consequences of losing any specific job and provide adequate support to workers while they search for employment. 

In particular, Congress should strengthen unemployment insurance protections and take steps to reduce the cost of health care and increase access to retirement security. These, along with other improvements to social insurance systems and workforce development programs, will not only ease the financial burden and pain of specific job transitions, but increase the bargaining power of all workers—especially those who aren’t represented by a union. The single greatest bargaining chip nonunion workers have is the ability to leave their jobs. Therefore, efforts that improve our social insurance systems and reduce workers’ reliance on employers for health care and retirement, among other benefits, will increase their ability to walk away from subpar working conditions, creating pressure on businesses to change their practices. 

Similarly, it is far better to lose one’s job when overall unemployment is very low than when it is high. Further, threats to leave an exploitative workplace and find a better job elsewhere are far more credible when the aggregate labor market is experiencing very low unemployment than when unemployment is high. Macroeconomic policy that targets sustained periods of very low unemployment and that quickly restores the labor market to health after recessionary shocks is vital for workers to carve out a decent career during periods of technological changes. Of all the policy failures that created the rise in inequality and the anemic wage growth in recent decades, macroeconomic policy failures likely top the list. Sound macroeconomic policy will dwarf any level of AI penetration in its importance to the trajectory of workers’ wages and employment over the next few decades.

Conclusion

Over the last few years, researchers, unions, and workers have drawn much needed attention to how the deployment of AI technologies has exacerbated harmful workplace conditions and outcomes for many workers. It’s understandable why policymakers are focused on developing legislation that addresses the tool—AI technology—used by businesses to achieve these outcomes. However, we believe that legislation aimed at improving outcomes for workers must also address the underlying conditions that enable businesses to deploy AI in this manner. Legislative interventions must carefully consider and address why workers don’t currently have the bargaining power to directly engage employers when they use AI—or any technology—in a manner that skirts their rights or creates poor working conditions. That is why we believe any serious worker-centered AI policy must restore core labor rights, strengthen social insurance programs, and maintain full employment. If these underlying economic conditions are not addressed, then we fear that legislative interventions that regulate how AI is developed or deployed will fall far short of their objectives.

In the coming months, we will release a series of follow-up pieces that will examine each pillar, laying out in more detail how Congress and states could design specific proposals to achieve the policy objectives of each pillar. We believe the three broad policy recommendations discussed above would most directly and effectively increase workers’ ability to meaningfully engage employers over the deployment of AI technologies in the workplace. The advancement of technology creates an opportunity for greater prosperity and economic growth. Under the right conditions, workers can and should fully share in those gains.

But we will sound a note of caution: It is possible for policymakers looking to help workers thrive in labor markets to focus too much attention on AI-specific issues. Simply put, there is no compelling evidence that AI is enormously different enough or more powerful than other earlier waves of technological change in its effect on labor markets. Further, the pre-AI status quo in labor markets was terrible for workers’ opportunities to thrive and for needed fundamental policy change across all sorts of policy areas that remain unaddressed. Given this fact, U.S. workers would not be well served by having policymakers, researchers, and advocates focus disproportionate amounts of attention on AI deployment at the expense of other crucially needed reforms. To put it more bluntly, it would serve the interests of exploitative employers to have the pro-worker policy community lose focus on many other issues to concentrate large amounts of attention and influence on AI specifically.

Appendix

The summary below aims to distill federal bills related to AI by highlighting the most common policy interventions and strategies. By focusing on specific types of interventions, we hope to highlight both the nuance of the policy strategies as well as how even the same form of intervention can be structured and implemented in a variety of ways.

Disclosure requirements

Across branches and levels of government, there are efforts to increase basic awareness among workers (and the public) of the presence of AI technologies used in the workplace.

The Stop Spying Bosses Act broadly requires employers to disclose—to employees and the public—their practices related to workplace surveillance and how these affect or influence employment-related decisions. More specifically, the legislation requires employers to disclose the following: what data are collected, how the data are being collected, where and when data are being collected, how frequently the collection occurs, what the business purpose of the collection is, and which, if any, third party service providers are engaged for the surveillance, data transfer, or sale of worker-level data, among other disclosures.6

Similarly, the No Robot Bosses Act generally addresses employment-related decisions that are informed by outputs from automated decision systems. The bill would require employers to provide employees with a plain-language explanation of outputs from automated decision systems including: a description of the system used, description and copy of the data inputs to the system, an explanation of how the outputs were used in making the employment-related decision, and the reason for using the automated decision system outputs in making the employment-related decision.7

The Algorithmic Accountability Act directs the Federal Trade Commission (FTC) to implement, through regulations, a new requirement for companies to assess and disclose the impact of automated decision systems, including those related to “employment, workers management, or self-employment.” Specifically, covered entities would be required to submit an initial summary report of their assessment to the FTC prior to the deployment of newly covered technologies; entities would also be required to submit annual reports on the ongoing assessment of technologies that are already on the market. The bill identities specific topics that should be disclosed in the summary reports to the FTC including the following: the purpose of the product and a detailed description of the decision(s) the system intends to make; any publicly stated guardrails or limitations on use of the product; documentation of the data used as inputs during the development, testing, and maintenance of the systems; and any transparency mechanisms that are included in the system to allow end users to contest, correct, or appeal decisions made by the system.8 

The Warehouse Worker Protection Act broadly aims to address productivity quotas that are opaque to employees, increase risk of injury, or are used as pretext for exercising labor rights; the bill includes a number of disclosure requirements. Specifically, the bill would require covered employers to provide a written description of each quota an employee is subject to. The description must be provided in plain language and include the following: the number of tasks that must be completed within a specified time period; what, if any, disciplinary actions could result from failure to meet quotas; how the employer measures work speed, including where and when measurements occur; and the businesses’ purpose for collecting work speed data. Employers that take adverse employment actions against workers as result of failing to meet established quotas would be required to provide employees with a written explanation of how the quota wasn’t met and a copy of work speed data.9 

Auditing and impact assessment requirements

As discussed above, one of the key labor market impacts of AI is the potential for technologies to undermine existing labor and employment rights, including anti-discrimination protections. A number of policy proposals and state-level laws attempts to address these issues by creating new pre-deployment and ongoing auditing requirements of technologies, particularly those used in hiring or other employment-related decisions. 

Under the No Robot Bosses Act, employers would not be able to use any automated decision system to inform employment-related decisions unless it underwent pre-deployment testing and validation. Specifically the bill requires validation with respect to compliance with employment laws including: Title VII, ADEA, ADA, FLSA, Rehabilitation Act, and Pregnant Workers Fairness Act, among other specifically identified protected classes. The proposed legislation would also require, at minimum, annual independent auditing to ensure continued compliance with these laws.10

Under the Algorithmic Accountability Act of 2023, companies that develop AI technologies would be required to conduct pre-deployment and ongoing impact assessment for automated decision systems. The bill identifies an extensive list of issues that developers should include in their assessments including: comparison of performance outcomes under test conditions and deployment conditions; evaluation of differential outcomes associated with race, color, sex, gender, age, disability, religion, family status, socioeconomic status, or veteran status; assessment of the need for guardrails or limitations on the use of products; and assessment of the explainability and transparency of systems for consumers.11

Prohibition of specific technologies or use cases 

Beyond increasing transparency through disclosure requirements and imposing new auditing requirements, a number of legislative proposals prohibits specific uses of AI technologies. In some instances, these prohibitions simply reiterate that certain actions are not allowed under existing laws; in other instances, the prohibitions attempt to curtail the use of AI technologies in ways that may be harmful to workers given their limited bargaining power and ability to alter employer practices on their own. 

Under the No Robot Bosses Act, employers would not be able to rely exclusively on automated decision systems to make hiring, firing, disciplinary, or leave-related decisions, though they could still use the systems as an input or factor in a decision-making process. This type of prohibition attempts to remove perhaps the most extreme use cases of the technology as it relates to employment decisions. Importantly, though, it is currently quite common for business groups and companies to claim that technology alone doesn’t make these decisions.12

Under the Stop Spying Bosses Act, employers would be prohibited from using surveillance technologies when covered workers are off duty or in sensitive areas such as a locker room or restroom. Additionally, the bill would prohibit the use of these technologies to identify individuals who are engaging in labor organizing activities— a use of technology which the NLRB general counsel has noted is, in her office’s view, unlawful under the NLRA. Additionally, the bill would prohibit employers from using surveillance data to identify workers’ political opinions, religious views, and health conditions and outcomes that are unrelated to the performance of job duties.13

The Warehouse Worker Protection Act includes provisions that would prohibit employers from requiring a quota that would prevent the following: compliance with any required meal or rest breaks; compliance with health and safety standards (required by federal, state, or local laws); employees’ use of bathroom facilities; and compliance with reasonable accommodations and nondiscrimination provisions of federal, state, and local laws. Additionally, the bill would prohibit employers from setting quotas that establish a performance target over a period that is less than a day and prohibits quotas that would prevent or discourage employees from exercising their rights under the National Labor Relations Act.14

Amending existing labor and employment laws; agency directives 

As discussed above, businesses’ use of AI technologies has exacerbated violations of existing labor and employment standards. Some legislation recognizes this by explicitly prohibiting technology to be used in a manner that violates these rights, such as nondiscrimination protections or labor organizing rights (see above). In addition to this intervention, some proposed legislation goes further to amend underlying statutes to bolster and expand protections, and direct agencies to issue regulations and reports. 

The Warehouse Worker Protection Act would notably amend the National Labor Relations Act to make it an unfair labor practice for employers to “impose on an employee a quota that significantly discourages or prevents, or is intended to significantly discourage or prevent, an employee from exercising the rights guaranteed in section 7.” Additionally, the bill would amend the NLRA to create a rebuttable presumption of discrimination if an employer takes an action to impose a quota against an employee within 90 days of exercising their rights under section 7 of the NLRA. Additionally, the bill includes a series of directives to OSHA regarding rulemaking and amends the OSH Act. Specifically, the bill would require OSHA to issue the following: a proposed ergonomics standard (within 3 years) and a proposed standard requiring covered employers to provide employees access to trained first aid professionals at the facility (within 1 year). Finally, the bill would amend the OSH Act to include provisions related to the correction of serious, willful, or repeated violations pending contest and procedures for a stay. 15

The Eliminating Bias in Algorithmic Systems Act generally requires every federal agency to establish an office of civil rights that is focused on combating AI bias and discrimination; the bill is not limited to agencies that regulate employment practices but is inclusive of them. While the bill does not direct agencies to issue any specific piece of regulation or even amend their underlying authorities, the bill directs agencies to engage in specific activities in order to better identify and reduce prevalence of bias and discrimination. Specifically, the bill requires agencies to submit a report to Congress on the state of the technology with respect to the jurisdiction of the agency and any relevant steps the agency has taken to mitigate algorithmic bias and discrimination.16

Workforce and training

As discussed above, one concern among policymakers is the impact of AI on job displacement and skills. There are a few bills that address this issue, though primarily by directing agencies to engage in further research. 

The Technology Workforce Framework Act generally expands the functions of NIST to include a workforce framework for emerging technologies. Specifically the bill would direct NIST to define AI-related jobs and the necessary knowledge, skills, and abilities needed to fill those jobs.17

The Jobs of the Future Act of 2023 directs the Department of Labor and the National Science Foundation to submit a report to Congress on AI and its impact on the workforce including: industries that are projected to have the most growth in the use of AI technologies and whether that use would result in job enhancement or job replacement; analysis of the skill necessary to develop or use AI technologies; methods to ensure necessary skills, expertise, and education are accessible to all segments of the current and future workforce; recommendations to minimize job displacement; and workforce training needs.18

Notes

1. See for examples: Ellingrud et al. 2023; Goldberg 2023; Kochhar 2023; Tamayo et al. 2023.

2. See for examples: Bernhardt, Suleiman, and Kresge 2021; Yang 2020.

3. See for examples: Khan 2024; Long 2022; Mims 2021.

4. See BLS 2023. In 2019 total nonfatal injury and illness rates per 100 full-time workers was 4.8; in 2022 (latest data available), the rate was 5.5.

5. Richard L. Trumka Protecting the Right to Organize Act of 2023, [H.R.20] 118th Cong. (2023) and Public Service Freedom to Negotiate Act of 2024, [S. 4363] 118th Cong. (2024).

6. Stop Spying Bosses Act, [S.262] 118th Cong. (2023).

7. No Robot Bosses Act, [S.2419] 118th Cong. (2023).

8. Algorithmic Accountability Act of 2023, [H.R.5628] 118th Cong. (2023).

9. Warehouse Worker Protection Act, [S.4260] 118th Cong. (2024).

10. No Robot Bosses Act, [S.2419] 118th Cong. (2023).

11. Algorithmic Accountability Act of 2023, [H.R.5628] 118th Cong. (2023).

12. No Robot Bosses Act, [S.2419] 118th Cong. (2023).

13. Stop Spying Bosses Act, [S.262] 118th Cong. (2023).

14. Warehouse Worker Protection Act, [S.4260] 118th Cong. (2024).

15. Warehouse Worker Protection Act, [S.4260] 118th Cong. (2024).

16. Eliminating Bias in Algorithmic Systems Act of 2023, [S.3478] 118th Cong. (2023).

17. Technology Workforce Framework Act of 2024 [S.3792] 118th Cong. (2024).

18. Jobs of the Future Act of 2023, [H.R.4498] 118th Cong. (2024).

References

Athena Coalition, National Employment Law Project, Warehouse Workers for Justice, Warehouse Worker Resource Center, United for Respect, Atwood Center, Make the Road New Jersey, Make the Road New York, and New York Communities for Change. 2024. Packaging Pain: Workplace Injuries in Amazon’s Empire, January 2020.

Bernhardt, Annette, Reem Suleiman, and Lisa Kresge. 2021. Data and Algorithms at Work: The Case for Worker Technology Rights. UC Berkeley Labor Center, November 2021.

Bivens, Josh, and Ben Zipperer. 2024. Unbalanced Labor Market Power Is What Makes Technology—Including AI—Threatening to Workers, Economic Policy Institute, March 2024.

Bureau of Labor Statistics (BLS). 2023. “Table 1. Incidence Rates of Nonfatal Occupational Injuries and Illnesses by Industry and Case Types, 2022,” Injuries, Illnesses, and Fatalities, 2022. Last modified November 8, 2023.

Center for Labor and a Just Economy Lab (CLJE Lab). 2024a. Building Worker Power in Cities and States: A Toolkit for State and Local Labor Policy Innovation. Harvard Law School, September 2024.

Center for Labor and a Just Economy Lab (CLJE Lab). 2024b. Worker Power and Voice in the AI Response. Harvard Law School, January 2024.

Chapman, Michelle, and Haleluya Hadero. 2024. “Amazon Labor Union Members Vote Overwhelmingly in Favor of an Affiliation with the Teamsters.” Associated Press, June 18, 2024.

Clean Slate for Worker Power (Clean Slate). 2021. “Principles of Sectoral Bargaining: A Reference Guide for Designing Federal, State, and Local Laws in the U.S.” Harvard Law School’s Labor and Worklife Program, May 2021.

Department of Labor (DOL). 2024. Memorandum, Subject: Artificial Intelligence and Automated Systems in the Workplace under the Fair Labor Standards Act and Other Federal Labor Standards. April 29, 2024.

Doellgast, Virginia, Sean O’Brady, Jeonghun Kim, Della Walters, Alyssa Acevedo, Nelson Dragsbaek, Tom Hegeman, Stefan Ivanovski, Natalie McCormick, Deepa Rajan, Ayaj Rana, MT Snyder, Jelena Starcevic, and Ines Wagner. 2023. AI in Contact Centers: Artificial Intelligence and Algorithmic Management in Frontline Service Workplaces. Cornell ILR School, November 2023.

Ellingrud, Kweilin, Saurabh Sanghvi, Gurneet Singh Dandona, Anu Madgavkar, Michael Chui, Olivia White, and Paige Hasebe. 2023. Generative AI and the Future of Work in America. McKinsey Global Institute, July 2023.

Equal Employment Opportunity Commission (EEOC). 2023. ” EEOC Releases New Resource on Artificial Intelligence and Title VII” (press release), May 18, 2023. 

Glass, Aurelia. “Unions Give Workers A Voice Over How AI Affects Their Jobs” Center for American Progress. May 2024.

Goldberg, Emma. 2023. “A.I.’s Threat to Jobs Prompts Question of Who Protects Workers.” New York Times, May 23, 2023.

Griffith, Kati L. 2019. “The Fair Labor Standards Act at 80: Everything Old Is New Again.” Cornell Law Review 104, no.557 (2019). https://scholarship.law.cornell.edu/clr/vol104/iss3/1.

Khan, Mishal. 2024. At Work and Under Watch: Surveillance and Suffering at Amazon and Walmart Warehouses. Oxfam, April 10, 2024.

Kochhar, Rakesh. 2023. Which U.S. Workers Are More Exposed to AI on Their Jobs? Pew Research Center, July 2023.

Lang Gordon, Julia. 2021. “Under Pressure: Addressing Warehouse Productivity Quotas and the Rise in Workplace Injuries.” Fordham Urban Law Journal, 2021 49, no. 1. https://ir.lawnet.fordham.edu/ulj/vol49/iss1/5.

Long, Katherine. “Amazon Workers Say Minor Aches Suddenly Became Debilitating as They Raced to Meet Speed Targets.” Business Insider, October 19, 2022.

Madland, David. 2019. How to Promote Sectoral Bargaining in the United States. Center for American Progress, July 2019.

Mims, Christopher. 2021. “The Way Amazon Uses Tech to Squeeze Performance Out of Workers Deserves Its Own Name: Bezosism.” Wall Street Journal, September 11, 2021.

National Labor Relations Board. 2022. “NLRB General Counsel Issues Memo on Unlawful Electronic Surveillance and Automated Management Practices” (news release), October 31, 2022.

National Nurses United (NNU). 2024. “National Nurses United Survey Finds A.I. Technology Degrades and Undermines Patient Safety” (press release), May 15, 2024.

O’Brady, Sean, and Virginia Doellgast. 2021. “Collective Voice and Worker Well-Being: Union Influence on Performance Monitoring and Emotional Exhaustion in Call Centers.” Industrial Relations: A Journal of Economy and Society 60, no. 3 (July): 307–337. https://doi.org/10.1111/irel.12286

Occupational Health and Safety Administration (OSHA). 2023a. “Department of Labor Announces National Emphasis Program Aimed at Reducing, Preventing Workplace Hazards in Warehouses, Distribution Centers” (news release). U.S. Department of Labor, July 13, 2023.

Occupational Health and Safety Administration (OSHA). 2023b. “OSHA nstruction: National Emphasis Program on Warehousing and Distribution Center Operations.” U.S. Department of Labor, July 13, 2023.

Occupational Safety and Health Administration (OSHA). 2024. “Ergonomics – Standards and Enforcement FAQs” (web page). U.S. Department of Labor. Accessed on September 4, 2024.

Rosenberg, Eli. “Fired Amazon Union Organizer at Kentucky Warehouse Alleges Retaliation.” NBC News August 25, 2022

Salaam, Ameenah. 2023. “Written Comments for AI Insight Forum on Workforce.” Washington, D.C., October 27, 2023.

Tamayo, Jorge, Leila Doumi, Sagar Goel, Orsolya Kovács-Ondrejkovic, and Raffaella Sadun. 2023. “Reskilling in the Age of AI.” Harvard Business Review, September 2023.

Tung, Irene, Nicole Marquez, and Paul K. Sonn. Amazon’s Outsized Role: The Injury Crisis in U.S. Warehouses and a Policy Roadmap to Protect Workers. National Employment Law Project, May 2024.

Yang, Jenny R. 2020. Three Ways AI Can Discriminate in Hiring and Three Ways Forward. Urban Institute, February 2020.

Tina T. Williams,
Director, Division of Policy and Program Development
Office of Federal Contract Compliance Programs
U.S. Department of Labor
200 Constitution Avenue NW, Room C-3325
Washington, DC 20210

Re: Comments on Supply and Service Program; Proposed Approval of Information Collection Requirements; FR Doc. 2022–25311

Dear Ms. Williams:

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. EPI submits these comments on the Office of Federal Contract Compliance Programs’ (OFCCP) request for reauthorization of its compliance review scheduling letter. EPI strongly supports the proposal that the scheduling letter collect more detailed and complete information at the outset of a compliance review.

A significant portion of the American workers stand to benefit from enhanced contractor compliance. OFCCP has jurisdiction over approximately 120,000 contractor establishments and 25,000 firms, which employ approximately 20% of the American workforce. And with the new historic federal investments for infrastructure and economic recovery, many more businesses will become federal contractors subject to OFCCP oversight. Enhancing OFCCP’s ability to make good jobs free from discrimination available to all is especially critical. This change is essential for OFCCP to conduct more efficient, consistent, and effective reviews of federal contractors’ compliance with nondiscrimination and equal employment opportunity requirements.

Despite long standing protections under the law, working people across the United States continue to experience employment discrimination that robs them of employment opportunities, economic security, and dignity on the job.1 The cost of discrimination for people of color, women, LGBTQ+ people, people with disabilities, veterans, and other marginalized and multi-marginalized groups is significant. Black workers, and Black women workers specifically, face the acute challenges of occupational segregation – overrepresentation of black workers, and in low-wage occupations and underrepresentation in higher-wage occupations2. Both workplace discrimination and occupational segregation can prevent access to a job or a promotion, cause a hostile working environment, or lower pay — all because of who you are. These unlawful practices inhibit economic security and opportunity and help to perpetuate disparities in health outcomes, housing, education, and more.

OFCCP is unique in being able to conduct systemic compliance reviews as part of its enforcement authority. Through compliance reviews, OFCCP can proactively identify, investigate, and remedy patterns of discrimination, even in the absence of an individual complaint, and can evaluate contractors’ compliance with affirmative action obligations. The scheduling letter, which OFCCP now proposes to revise, is the document OFCCP uses to notify contractors that they have been selected to undergo a compliance review and identifies the initial information those contractors must provide.

OFCCP proposes that its scheduling letter request more detailed and specific information from contractors at the outset of compliance reviews. Updating the scheduling letter to obtain critical information at the beginning of the compliance review will support OFCCP’s goal of strengthening the effectiveness of its compliance evaluations, promoting greater contractor compliance, and ultimately benefiting more workers. It will also encourage employers to self audit the employment systems referenced in OFCCP’s updated requests (e.g., technology-based employment systems) to identify potential EEO issues before they are selected for a compliance review by the OFCCP. The new information would include:

• Existing employment policies concerning equal opportunity, including anti-harassment policies, EEO complaint procedures, and employment agreements, such as arbitration agreements, that impact employees’ equal opportunity rights and complaint processes. Having this information at the outset is essential for OFCCP to understand the contractors’ systems and proceed with an informed and targeted review.
• More details about the number of qualified people of color and women available for employment in each job group, enhancing OFCCP’s ability to evaluate contractors’ affirmative action programs.
• More detailed information about promotions and terminations, including information necessary to make the review meaningful, such as whether the promotions were competitive and the reason for termination.
• New information on the contractor’s use of technology-based employment selection procedures, including artificial intelligence, algorithms, and automated systems, made essential given the documented potential for bias in such systems.3

If authorized as proposed, these changes will speed the pace of reviews, conserve scarce agency resources, provide additional clarity for employers as to their obligations, and enable OFCCP to more quickly and accurately identify both potential problem areas and successes.

OFCCP has tailored the proposed changes to the scheduling letter to minimize the additional burden on contractors. As the agency explains, the new scheduling letter would reduce contractor uncertainty over what documentation is sufficient for the review and enhance review efficiency for both the contractor and OFCCP. For these reasons, EPI strongly supports the proposed reauthorization of its compliance review scheduling letter.

Adewale Maye
Research and Policy Analyst
Program on Race, Ethnicity, and the Economy
Economic Policy Institute

1. See, e.g., U.S. Equal Employment Opportunity Commission, Enforcement and Litigation Statistics, https://www.eeoc.gov/data/enforcement-and-litigation-statistics-0; Office of Federal Contract Compliance Programs, OFCCP By the Numbers, available at https://www.dol.gov/agencies/ofccp/about/data/accomplishments.

2. Valerie Wilson and William Darity Jr, Understanding black-white disparities in labor market outcomes requires models that account for persistent discrimination and unequal bargaining power, Economic Policy Institute, 2022.

3. Manish Raghavan & Solon Barocas, Challenges for mitigating bias in algorithmic hiring, Brookings (Dec. 6, 2019), https://www.brookings.edu/research/challenges-for-mitigating-bias-in-algorithmic-hiring/ (“Left unchecked, algorithms can perpetuate the same biases and discrimination present in existing hiring practices.”); Miranda Bogen, All the Ways Hiring Algorithms Can Introduce Bias, Harvard Business Review (May 6, 2019), https://hbr.org/2019/05/all-the-ways-hiring-algorithms-can-introduce-bias.

The growing evidence that high-pressure labor markets can narrow these gaps calls for macroeconomic policies that test the absolute limits of how low unemployment can be pushed. Macroeconomic policymakers frequently weigh the potential benefits of higher-pressure labor markets against the potential risks, namely accelerating price inflation driven by excessively fast wage growth. The potential to close race-based gaps in the labor market should be counted as a substantial benefit in these deliberations and should convince policymakers to take on more inflation risk. Furthermore, running labor markets at the maximum sustainable pressure will provide much-needed information on what else we need to do to foster racial equity in labor markets.

This paper explores the promise and limits of high-pressure labor markets in reducing racial labor market gaps and how too-slack labor markets have helped thwart progress in closing the gaps. It then draws lessons from these investigations for policymakers. Its main findings are:

• Reductions in the unemployment rate boost hourly wages of typical (median) Black workers more than they boost hourly wages of typical white workers.
  • In 2019, the median Black worker was paid 32.2% less in hourly wages than the median white worker, up from 28.6% in 1973. Had the unemployment rate averaged 1 percentage point less annually from 1973 to 2019, the median Black–white wage gap could have declined to 18.0%. If the unemployment rate had averaged 2 percentage points less (a very ambitious target), the median Black–white wage gap could have fallen to just 5.4% (an 80% reduction in the size of this wage gap).
• Reductions in the unemployment rate provide an even bigger relative boost for median Black annual earnings, by increasing both hours worked and hourly wages.
  • In 2019, the annual earnings of the typical (median) Black worker amounted to just 80% of the annual earnings of the median white worker. Had the unemployment rate averaged 2 percentage points less annually over the 1970–2019 period, the Black–white median earnings gap (measured as a ratio) could have essentially closed. Had the unemployment rate averaged 1 percentage point less annually over that period, the typical Black worker in 2019 could have been paid 90.1% as much as the typical white worker—reflecting a 50% decrease in the Black–white median annual earnings gap.
• The Black–white unemployment gap (how much, in percentage points, the Black unemployment rate exceeds the white unemployment rate) closes significantly when the overall economy has fewer idle resources (i.e., when potential output climbs closer to actual output, leading to a rise in the measured “output gap”). For example, the Black unemployment rate falls more than twice as much as white unemployment when the economy’s output gap rises by 1 percentage point.
  • Even this disproportionate reduction in Black unemployment might understate how equalizing overall economic tightening can be. When the output gap measure rises 1 percentage point, the share of Black persons who are employed (the Black employment-to-population ratio, or EPOP) actually rises nearly seven times as much as the share of white persons employed (the white EPOP). But because the share of Black persons who are either working or actively looking for work (the Black labor force participation rate) also rises faster than white labor force participation when the output gap improves, the Black unemployment rate reduction is muted relative to gains in employment.
• Sustained high-pressure labor markets may have more power than we thought to close the Black–white unemployment ratio. For decades, the Black unemployment rate has been, on average, roughly twice the white unemployment rate. This persistent and distressingly high Black–white unemployment ratio (the Black unemployment rate divided by the white unemployment rate) has traditionally been seen as much more resistant to closing with high-pressure labor markets. However, the Black unemployment rate has only been included in most data sets since the early 1970s and, since then, genuinely high-pressure labor markets have been quite rare. Pre-1970s data that provide a potential proxy for the Black–white unemployment ratio show that sustained high-pressure labor markets may well actually reduce it significantly.

The policy lessons from this data are clear. While overall wage and price inflation remain the proper targets of policy (employment measures are not reliable enough to make good policy guides), policymakers need to change how they balance those targets:

• As they weigh the potential benefits of higher-pressure labor markets against the risks, policymakers should count, on the benefits side, potential reductions in chronic racial gaps in labor market outcomes.
• More forbearance should be exercised as wages and prices rise during economic recoveries and expansions, and at a bare minimum wage and price targets should be kept symmetric over business cycles: Every year that sees wage and price inflation come in 1% below target must be matched by a year with wage and price inflation coming in 1% above target.
• The potential to close race-based gaps in the labor market should convince policymakers to take on more inflation risk than they otherwise would have (that is, they should wait for actual and sustained, rather than forecast, inflation to appear before raising interest rates).

Background on the unemployment and inflation trade-off

All else equal, policymakers should aim for an unemployment rate so low that it reflects only the transitory and voluntary shifts of workers in and out of work or between employers. However, because of the way policymakers have traditionally sought to affect the rate of unemployment, they have instead aimed for a rate that was high enough to avoid any chance, even remote, of sparking inflation.

The primary way policymakers influence the unemployment rate is through measures that change the pace of aggregate demand growth. Aggregate demand is economywide spending of households, businesses, and governments. When this spending is strong, employers need workers to produce the output of goods and services needed to satisfy customer demand, which keeps unemployment low and employment growth strong. When this spending lags, less output and hence fewer workers are needed to satisfy demand, so employment growth lags and unemployment rises.

If policymakers boost economywide spending too much, however, demand might outstrip the productive capacities of firms. As demand runs ahead of supply, this puts upward pressure on wages and prices as firms scrambling to meet demand find they need to hire more workers and can charge customers a bit more for scarce goods. This “inflation barrier” to further efforts to boost demand—the point of tightness in labor markets that sparks an upward drift of inflation—may well be hit before the unemployment rate that reflects only voluntary job transitions is attained.

This balancing between demand growth that is strong enough to keep unemployment low, but not strong enough to generate accelerating inflation, is a central problem of macroeconomic policy (often called stabilization policy). Traditionally, the entity doing this balancing in the United States has almost always been the Federal Reserve, which tries to spur demand primarily by lowering interest rates and can brake escalating demand by raising interest rates. However, the Great Recession exposed the extreme limits of the Fed’s ability to generate strong enough demand growth and has elevated the role of fiscal policymakers (Congress and the president) in boosting (or failing to boost) demand by adjusting spending levels and taxation in the economy.1

Far too often in recent decades, policymakers have erred in targeting—or at least unnecessarily tolerating—demand growth that was too weak to generate enough pressure in labor markets to give workers leverage in wage negotiations with employers.2 This toleration of low-pressure labor markets was often done in the name of keeping inflationary pressures in check. But given that genuine inflationary pressures in the U.S. economy have been extraordinarily rare since the 1970s, the targeting of too-weak demand growth has often been about guarding against even the risk of inflation. A growing body of recent research notes that the benefits of low unemployment are large enough to justify taking on substantially more inflation risk than has previously been tolerated.

The most obvious benefits of low unemployment are more job opportunities for more people and more hours of work available to U.S. families. A less obvious benefit, but one that shows up strongly in the data, is faster hourly wage growth for the vast majority of U.S. workers, a particularly important benefit given the anemic pace of wage growth for these workers in recent decades.3 Yet another increasingly discussed benefit of low unemployment is its ability to put sustained pressure on compressing race-based gaps in the labor market. The rest of this paper largely tries to put some empirical bounds on just how large this last benefit might be.

Why aim for ‘high-pressure’ labor markets and not ‘full employment’?

The Fed’s legal mandate is to pursue maximum employment consistent with price stability. Over the years “maximum employment” has often been referred to as “full employment.” However, there is no universally agreed upon definition of full employment. For some, full employment simply means that anybody who wants a job can find a job. For others, particularly macroeconomists, it means attaining the rate of unemployment (often called “the natural rate”) below which further increases in economywide spending will mostly lead to accelerating inflation rather than greater output. “High-pressure” labor markets just mean labor markets characterized by low unemployment, fast rates of job creation, rapid job-finding among the unemployed, and sustained effort by employers to keep their enterprises properly staffed. Labor markets can be “high pressure” yet still tolerate further reductions in unemployment without leading to unsustainable wage or price inflation, and the term “high-pressure” may better connote a continuum of labor market states rather than a single fixed point.

Further, old theories of “disguised unemployment” and new developments in advanced capitalist economies (the rise of “gig work”) argue that “full employment” and “high-pressure labor markets” might not always coincide.

Joan Robinson (1936) defined “disguised unemployment” as follows:

In a society in which there is no regular system of unemployment benefit, and in which poor relief is either nonexistent or “less eligible” than almost any alternative short of suicide, a man who is thrown out of work must scratch up a living somehow or other by means of his own efforts. And under any system in which complete idleness is not a statutory condition for drawing the dole, a man who cannot find a regular job will naturally employ his time as usefully as he may. Thus, except under peculiar conditions, a decline in effective demand which reduces the amount of employment offered in the general run of industries will not lead to “unemployment” in the sense of complete idleness, but will rather drive workers into a number of occupations—selling match-boxes in the Strand, cutting brushwood in the jungles, digging potatoes on allotments—[that] are still open to them.

The modern U.S. economy obviously does not totally lack relief for the unemployed, and the reach and influence of the gig economy is often wildly overstated. But it seems clear that one margin of survival that many U.S. workers draw on when regular work is slack due to weak aggregate demand is to engage in gig or otherwise irregular work. But gig work generally does not provide high-quality jobs or economic security. In some deeply unsatisfactory sense, the rise of gig work could theoretically help fulfill the promise of one definition of full employment—that anybody “who wants a job can find a job.” But, in an economy with measured unemployment kept low only by a large incidence of gig work, if policymakers boosted aggregate demand, it is highly likely that many gig workers would leave the gigs behind and look for and find more regular work. In short, describing labor markets as “high pressure” might better describe the condition that employers are competing actively among themselves to attract workers.

High-pressure labor markets and median racial wage gaps

Since 1979, wage gaps between Black and white workers have widened significantly. Figure A shows the gap in two ways: how much less in percent terms the median Black worker earns in hourly wages than the median white worker, and the percent by which the average hourly wage of Black workers is less than the average hourly wage of white workers, holding other characteristics constant. The latter, a regression-adjusted average gap, controls for educational attainment, gender, ethnicity, and age. Both gaps widened significantly over time, but the median gap started larger and has expanded more rapidly since the late 1970s.

Figure A

Previous work has indicated that median Black wage growth responds more strongly to changes in unemployment than does median white wage growth.4 Figure B confirms this. The figure shows the relationship between the unemployment rate and wage growth. Specifically, it shows the change in wage growth that occurs if the unemployment rate rises by 1 percentage point. For white median hourly wages, a 1-percentage-point increase in overall unemployment is associated with wage growth that is 0.52% slower. For Black median wages, wage growth declines by 0.76%. As the figure shows, the coefficient for median Black wage growth is nearly 50% larger than for median white wages.5

Figure B

Figure C uses the estimated coefficients from Figure B and calculates counterfactual median Black–white wage gaps under three scenarios: unemployment rates that averaged 1, 1.5, and 2 percentage points lower over the 1973–2019 period. These scenarios are plausible alternatives of what might have been under a policy regime that determinedly aimed for high-pressure labor markets. Over this period, the unemployment rate was high, averaging 6.2%. One closely watched measure of the unemployment rate consistent with stable inflation—the nonaccelerating inflation rate of unemployment (or NAIRU) estimated by the Congressional Budget Office (CBO)—averaged 5.3% over this same period, almost a full percentage point lower. Additionally, between 1947 and 1973, the unemployment rate averaged 4.7%, exactly 1.5 percentage points lower than in the post-1973 period, and inflation before the oil price shock of 1973 was generally contained. Finally, when unemployment fell more than 2 percentage points beneath the 1973–2019 average in the late 1990s, and again in 2018–2019, there was no marked uptick in wage or price inflation requiring that macroeconomic policymakers slow demand growth.

Figure C

We should be clear that examining a counterfactual that assumes substantially lower unemployment on average does not reflect an assumption that recessions never happen. Instead, it simply assumes that macroeconomic policymakers do their job and ensure that every period of above-average unemployment is matched by an equivalent period of below-average unemployment. Estimates of the natural rate of unemployment are not hard floors below which the economy is never meant to go; instead they are averages that the unemployment rate should fluctuate both above and below. Running the economy far above even too-conservative natural rate estimates over decades is a policy failure that can clearly be addressed.

Achieving and sustaining high-pressure labor markets since the early 1970s would have dramatically narrowed the median Black–white wage gap. Had unemployment averaged 2 percentage points less over the entire period, 80% of the median Black–white wage gap that appeared in 1973 could have been erased (as the gap shrank from 28.6% to 5.4%). With unemployment averaging just 1 percentage point less (essentially just hitting conventional measures of the natural rate of unemployment), the median wage gap could have fallen slightly (to 18.0%) rather than rising by almost 8 percentage points over this period. In short, high-pressure labor markets hold great potential to reduce this particular measure of racial inequality in the labor market.

The gap-narrowing power of high-pressure labor markets is even more evident when looking at median annual earnings. Annual earnings can be affected by tighter labor markets not only through higher hourly wages but also through increased hours worked during the year. As shown in Figure B, the decline in Black worker annual earnings associated with an uptick in the unemployment rate is an even larger decline than the decline in Black worker hourly earnings. Applying the same counterfactual scenarios of unemployment rates that average 1, 1.5, and 2 percentage points lower over the 1973–2019 period yields dramatic results for the median Black–white annual earnings gaps, shown in Figure D. In this figure, the gaps are presented as ratios—how much Black workers earn as a share of what white workers earn.

Figure D

Between 1970 and 2019, the ratio of Black to white annual earnings rose from 62.4% to 80.0%. If the unemployment rate had averaged 1 percentage point lower after 1973, then this ratio could have surpassed 90% by 2019. If the unemployment rate had averaged 2 percentage points lower, the Black–white annual earnings ratio would have essentially been 1, indicating near-complete equality in this measure.

High-pressure labor markets and gaps in employment and unemployment

As we have frequently noted, the Black unemployment rate has been, on average, roughly twice the white unemployment rate since 1972 (the first year that Black unemployment is measured by the Bureau of Labor Statistics). Further, as shown in Figure E, this rough 2-to-1 ratio prevails if one looks at the measure of “nonwhite” unemployment compiled by the BLS before 1972 (Black workers accounted for a very large majority of nonwhite workers over that pre-1972 period).

Figure E

The Black–white unemployment ratio shrinks only slightly if one adjusts for age, educational credentials, and gender composition of the workforces. In the figure, this ratio is calculated by comparing the “Black adjusted” line (see figure note) with the line for the white unemployment rate. For example, between 1976 and 2019, the overall Black–white unemployment ratio averaged 2.3 while the adjusted ratio averaged 2.0. This is an improvement for sure, but a depressingly small one, at roughly 15%.

One conclusion that can be drawn from Figure E is that the ratio of Black to white unemployment is pretty stubborn: It does not seem to fall quickly during periods of labor market tightness (when all rates fall together). However, even if the Black-to-white unemployment ratio never moved, the raw gap in unemployment rates between Black and white workers (simply the Black unemployment rate minus the white unemployment rate) would shrink rapidly during periods of overall labor market tightness, and would expand rapidly during periods of overall labor market distress. At a minimum, this means that Black workers see disproportionate gains and losses from effective and ineffective macroeconomic stabilization policy, respectively. Thus, getting macroeconomic stabilization policy right is a key issue for racial equity.

Figure F confirms this intuition, using the output gap as a proxy for overall economic, and thus labor market, health.6 The output gap is a measure of how fully the economy’s resources are being utilized at any given point in time (resources including potential workers). Specifically, it is calculated as the quotient of actual gross domestic product (GDP) divided by a measure of potential GDP (what GDP could have been had the economy’s resources been fully utilized), minus 1. When actual GDP is lower than potential GDP, the output gap is negative. As actual GDP falls further and further behind potential GDP, the gap measure becomes more negative; as it comes closer to potential GDP, it rises.

Figure F

Given this, as the gap rises, resource utilization increases and the overall unemployment rate generally falls. The figure shows the relationship between the overall output gap and Black and white labor market indicators. As can be seen, the Black unemployment rate is twice as responsive as the white unemployment rate to a change in the output gap: specifically, a 1-percentage-point increase in the output gap (moving actual GDP 1% closer to potential GDP) is associated with a 1.57-percentage-point decline in the Black unemployment rate, compared with a 0.64-percentage-point reduction in the white unemployment rate. Because the BLS measures the unemployment rate of Black workers only after 1971, we also include a measure of the responsiveness of what the BLS labels “nonwhite” unemployment—a series that goes back to 1954. In the years before 1972, Black workers made up more than 90% of those labeled nonwhite.7 The advantage of using a series with a longer historical perspective is that one can examine a period of very tight labor markets that were achieved in the mid-to-late 1960s. The overall unemployment rate, for example, fell to under 4% for three straight years in the late 1960s. In this longer time series, the overall responsiveness of the nonwhite unemployment rate to an increase in the output gap measure (-1.54) is quite close to the responsiveness of Black unemployment in the more recent series.

These differential rates of responsiveness translate into substantial closing of the gap between Black and white unemployment rates when the overall economy tightens up, with this gap defined simply as the Black unemployment rate minus the white unemployment rate (i.e., the gap is by how many percentage points the Black unemployment rate exceeds the white unemployment rate). The figure shows that each percentage-point increase in the output gap (i.e., a tightening of the economy and labor market generally) is associated with a 0.70-percentage-point reduction in the Black–white unemployment gap. Each percentage-point increase in the output gap is also associated with a 0.90-percentage-point reduction in the nonwhite–white unemployment gap. It is possible that relatively greater responsiveness of the nonwhite–white unemployment gap is due to the inclusion of workers who are not Black in the nonwhite unemployment calculation. It is also possible that the difference is due to the longer time series available with the nonwhite–white unemployment rate gap. By restricting this series to just post-1971 data points (to make it consistent with the Black unemployment rate coverage), the responsiveness of the nonwhite–white unemployment gap to a 1-percentage-point reduction in the output gap shrinks to 0.83 percentage points.

Race-based differentials in the responsiveness of labor market indicators to a change in the output gap are even larger when examining the responsiveness of the Black and white employment-to-population ratios. The Black EPOP rises by 1.2 percentage points as the output gap increases, while the white EPOP rises by 0.18 percentage points, just over a seventh as much.

If the ratio of Black to white unemployment rates was constant, then the change in the gap between these rates would also just equal this ratio multiplied by the change in the white unemployment rate. Given the relative stubbornness of the Black–white unemployment ratio (for example, as seen in Figure E), it might seem that it is essentially constant regardless of the state of labor market pressure. But it may not be.

Can high-pressure labor markets reduce Black–white unemployment ratios, not just gaps?

Looking at the Black and white unemployment rates over time—like those displayed in Figure E—can easily convince observers that the ratio of Black to white unemployment is nearly constant. In good times and in bad, the Black unemployment rate looks to be roughly twice the white unemployment rate. But there are actually some reasons for optimism—tempered, to be sure—that this ratio is not as unyielding to change as it seems. For one, there seems to be a shallow but steady downward trend in this ratio over time. For another, more detailed evidence indicates that the Black–white unemployment ratio may indeed respond measurably to high-pressure labor markets. That evidence is highlighted in the discussion of the next two figures, which show the overall unemployment rate and the Black–white unemployment ratio (Figure G) and the nonwhite–white unemployment ratio (Figure H) prevailing at business cycle peaks. Both show a clear positive relationship between the overall unemployment rate and the respective ratios (i.e., an increase in one measure coincides with an increase in the other). All else equal, this would indicate that a higher-pressure labor market overall does indeed put downward pressure on the Black–white unemployment ratio.

Figure G

Figure H

However, the positive relationship in Figure G is likely driven in some part by the trend in the Black–white unemployment ratio coinciding with the fact that, since 1976, later business cycles have consistently achieved lower unemployment rates. Between the business cycle peaks of 1979 and 2019, the Black–white unemployment rate ratio actually declined by a bit over 20% (as did the adjusted ratio, a ratio that estimates what the Black unemployment rate would have been had the composition of the Black labor force shared the same age, educational credentials, and gender mix as the white labor force). This trend likely would have led to successively lower Black–white unemployment ratios in 1989, 2000, and 2019 anyhow. But on top of this trend, unemployment rates in these business cycle years were successively lower over time. Given this, it is not clear if it is a given year’s unemployment rate or a long-running trend that drives the pattern in Figure G.

To test the connection, Figure H includes data on nonwhite unemployment back to 1959 and thus includes business cycles not characterized by uniformly lower overall unemployment rates over time. The strong positive relationship between rising overall unemployment and an increasing Black–white unemployment rate ratio still holds.

Figure I looks at the responsiveness of various labor market ratios (not gaps, as was analyzed above in Figure F) to changes in the output gap while controlling for a time trend. The first three data points come from a regression that used a lagged measure of the output gap. They show a significant decline in the Black–white and the nonwhite–white unemployment rate ratios associated with each percentage-point increase in the output gap (remember, as the economy improves and actual GDP gets closer and closer to potential GDP, the output gap rises). As before, this analysis includes a look at the coefficient on the nonwhite–white unemployment ratio from this regression just in the years after 1971 to see if some of the difference between its responsiveness and the responsiveness of the Black–white unemployment ratio is simply due to different timespans. The responsiveness of the nonwhite–white unemployment ratio is roughly same (but actually increases slightly) when just looking at the post-1971 period.

Figure I

The results from estimating the responsiveness of ratios of EPOPs and labor force participation rates (LFPRs) demonstrates a similar pattern as that shown by Black–white employment gaps. When the output gap rises, the Black–white ratio of EPOPs rises, meaning that the share of Black persons employed approaches closer to the share of white persons employed. A similar increase holds for the ratio of nonwhite-to-white EPOPs and, again, the responsiveness of this ratio falls a bit when only the post-1971 period is examined. The ratio of Black to white LFPR rises when output gaps rise as well, meaning Black labor force participation approaches closer to white labor force participation. Again, the responsiveness of the ratio of nonwhite–white LFPRs is greater than for the Black–white ratios, but much of this seems due to the different time periods; when only post-1971 years are included, the responsiveness is very similar.

The upshot of this examination is that there is some suggestive evidence that even key labor market ratios (not just absolute gaps) that compare labor market performance for Black and white workers might indeed narrow during periods of high-pressure labor markets. This more hopeful interpretation may have been missed by those looking only at raw measures of the Black and white unemployment rates over time (like those shown in Figure E).

First, the post-1971 period might not contain enough episodes of truly tight labor markets to allow the relationships between high-pressure labor markets and the Black–white unemployment ratio to be well estimated. Figure I has some slight suggestive evidence of this: The responsiveness of the nonwhite–white EPOP and LFPR is greater when the pre-1971 period is included. This pre-1971 period includes a long stretch in the 1960s when unemployment was beneath 4% for four straight years (1966–1969). Further, Aizer et al. (2020) found that many racial gaps in labor markets (like the Black–white earnings gap and the measures of occupation segregation) fell significantly in the 1940s. These declines were concentrated in areas with more defense spending. This spending not only contributed to tighter labor markets but also often came attached with anti-discrimination conditions. The authors could not disentangle the precise effect of each of these influences, but the large spillovers of reduced race-based gaps in industries not directly affected by the defense spending suggests a large role for high-pressure labor markets generally. And the U.S. labor market of the 1940s was high pressure in a way not seen since: The unemployment rate was below 2% for three straight years between 1943 and 1945. In short, since we have begun measuring the Black unemployment rate specifically, we just may not have seen enough periods of genuinely high-pressure labor markets to get a solid statistical read on what happens to the Black–white unemployment ratio when labor markets get truly tight and are kept that way for a sustained period of time.

Second, while the employment rate of Black workers rises significantly faster than for white workers as the output gap rises, the labor force participation rate of Black workers also rises faster, muting any disproportionate decline in unemployment for Black workers. If the measure is simple joblessness and not unemployment, then it seems clear that the Black–white jobless ratio clearly declines when labor markets tighten up.

Part of the reason why the stronger responsiveness of the Black–white EPOP to output gap increases translates weakly into a reduction in the Black–white unemployment ratio is likely due to different dynamics of labor force participation over the business cycle. A recent paper by Cajner, Coglianese, and Montes (2020) makes a significant contribution in our understanding of the cyclical behavior of labor force participation. Their main finding is that the LFPR is indeed affected by the state of labor market tightness, but that it responds substantially more slowly to positive or negative shocks than employment or unemployment. They further find that the Black LFPR responds substantially more strongly to a negative shock to the overall labor market. So when economic growth slows and the labor market develops slack, the rise in the Black unemployment rate relative to the white unemployment rate can be somewhat muted because of a larger labor supply response from Black workers. This means that the Black–white unemployment ratio may actually fall during recessions. Just to buttress this point, it is striking that the two lowest annual Black–white unemployment ratios on record occurred in 2009 and 2020—two of the worst years for economywide labor market health in the past 70 years or more.

As recoveries begin, Black EPOPs respond more strongly to improving economic conditions. All else equal, this should lead to a reduction in the Black–white unemployment ratio. But because the Black LFPR recovers more quickly than the white LFPR , the progress in reducing the racial unemployment gap is blocked by faster labor force growth among Black workers. By the time late in recoveries when labor markets are getting tight again, the white LFPR likely begins recovering more strongly, which allows for a reduction in the Black–white unemployment ratio. This modestly complicated series of dynamics likely explains part of why the salutary effect of lower unemployment rates on the Black–white unemployment ratio might be harder to detect in a simple eyeballing of trends. In 2019, the last business cycle peak, the Black–white unemployment ratio hit its lowest point at any business cycle peak on record. This likely reflects both a shallow but nontrivial downward trend over time and pressure that tight labor markets put on compressing the ratio. Additionally, the prolonged (if too slow) recovery following the Great Recession allowed ample time for the white LFPR to recover from the negative shock of the Great Recession and to stop putting downward pressure on the white unemployment rate.

Policy implications

The upshot of this examination is that sustained periods of high pressure in U.S. labor markets might significantly narrow racial gaps in unemployment and other key labor market measures. Given the long history of structural racism in the United States and the intentional policy efforts that created these gaps, it seems incumbent upon policymakers to use every tool available to try to close them. High-pressure labor markets look as promising as (or more promising than) any other tool. The large benefits—moral, political, and economic—of closing these labor market gaps call upon macroeconomic policymakers to consider them when assessing the benefits and costs of a “go for growth” strategy targeting high-pressure labor markets. To be explicit: The potential of more aggressive expansionary macroeconomic policy to help close race-based gaps in the labor market is worth taking on more risk of sparking inflation.

This policy recommendation for macroeconomic policymakers (including the Federal Reserve) to take on extra inflation risk in the name of narrowing racial gaps in the labor market is likely frustratingly imprecise to some. Some policymakers would prefer the clarity of, say, a numerical target for the Black unemployment rate. However, excess confidence in the ability of macroeconomic policymakers to use hard-and-fast ex ante labor market targets that precisely define “high pressure” has backfired in the past. Specifically, that unfounded confidence is a prime reason why labor markets were kept too slack for so long in recent decades, as hard targets such as estimates of the NAIRU turned out to be wrong, leading to unemployment rates in excess of what was needed for reasonable inflation control. Further, if using overall unemployment rates as precise labor market targets has proven to lead to unsatisfactory outcomes (and it has), using the Black unemployment rate as a specific target might be even worse, as one would be implicitly targeting not only the overall rate, but also how robustly the ratio between the Black and the overall rate changed as overall unemployment rose and fell depending on labor market conditions.

One of the most direct and thoughtful calls for having the Federal Reserve aim for narrower racial gaps in the labor market was by Bernstein and Jones (2020b). Their paper is often described as calling on the Fed to “target the Black unemployment rate,” but it does so only in the sense described above: It calls upon the Fed to consider the benefits of narrower gaps, and explicitly make them part of their criteria for decision-making. As the authors explain, “It is not just asking the chair to tell us about the gaps; it requires him or her to make closing them a part of their mandate” (Bernstein and Jones 2020a).

These sensible calls to narrow labor market gaps do raise an important question: Why is there reticence to demand that macroeconomic policymakers achieve a full elimination of labor market gaps? The answer is because it is unlikely that macroeconomic policy by itself can neutralize the centuries-long legacy of structural racism. This legacy has led to disadvantages Black workers face along numerous margins in the labor market, and while high-pressure labor markets can help to ameliorate these disadvantages, high-pressure labor markets likely cannot completely undo them before inflationary pressures require some moderating of expansionary policy.

For example, some of the gap in unemployment rates between Black and white workers represents differing levels of educational credentials. As we showed earlier (Figure E), adjusting the Black unemployment rate under a scenario that gives the Black and white workforces the same age and educational profiles does reduce the Black–white unemployment ratio by a small amount, around 15%. This gap in educational credentials obtained by Black and white workers is itself largely a function of historic discrimination, but it is unlikely to be solved simply by boosting aggregate demand. Further, even at the same level of educational credentials, it is certainly possible for the quality of educational investments to differ systematically between Black and white workers. Research has shown that educational investments are not only larger in white neighborhoods, but they have also been systematically reduced in schools with larger shares of Black students.8 Thus it seems likely that equalizing labor market outcomes will require interventions over and above expansionary macroeconomic policy to address the differences in education investment.

Despite these caveats, the results in this paper and previous research clearly show that expansionary macroeconomic policy can have profoundly equalizing effects. It also seems clear that policymakers have not fully accounted for these benefits when weighing benefits against the potential cost of sparking inflationary pressure. Further, ignoring these potential benefits may even result in worse analytical forecasting. Concepts like the natural rate of unemployment and the level of potential output for the U.S. economy often are estimated by assuming a given Black–white unemployment gap that does not close as the economy heats up.

An oft-cited example is the Congressional Budget Office (CBO) estimate of the natural rate of unemployment. The CBO assumes the overall unemployment rate reached in 2005 is consistent with the economy’s natural rate. It then takes group-specific unemployment rates that prevailed in 2005 and allows the overall natural rate to change only as the group-specific shares of the labor force change over time due to demography or immigration flows (Shackleton 2018). In some sense, this method implicitly assumes that group differences in unemployment that prevailed in 2005 are set in stone. Some have gone so far as to call this assumption racist. This seems wrong. The existence of the gaps is evidence of racism. But it would be odd indeed to ignore them entirely when doing forecasting given how persistent they have been. Instead, it seems that this assumption of ever-persisting gaps is evidence more of pessimism (much of it arguably well-earned) than of racism. 

But if these gaps do indeed close further as labor markets enter high-pressure periods, then any estimate of the natural rate of unemployment can be lower and estimates of potential output can be higher than one would otherwise forecast. In essence, we will never know the full extent of other policy interventions that need to be done to foster racial equity until we have fully maximized the reach of high-pressure labor markets. To put this another way, we won’t even know the size of the Black–white unemployment gap until we are sure we have reached the lowest rate of overall unemployment consistent with sustainable inflation. And yet for the vast majority of years over recent decades, we have not been close to this minimum unemployment rate.

In recent months, handwringing about the possibility of eventually “overheating” the U.S. economy due to excessively generous fiscal stimulus has begun. It is true that we are not completely certain about how low unemployment can go or how much the economy’s supply side will respond to growth in aggregate demand—so signs of overheating should indeed be monitored. But we should be very cautious about premature declarations of overheating. We have not seen sustained and broad-based wage and price inflation in the U.S. economy for decades. And we now know much more than in previous years about just how equalizing a high-pressure labor market can be, both for compressing wage growth among low-, middle-, and high-wage workers and for closing race-based gaps in labor market measures. These benefits are utterly enormous, and maximizing them is worth the risk of being very patient before aiming to deflate high-pressure labor markets through policy.

Endnotes

1. See Bivens 2016 for the central role of fiscal policy in conditioning economic growth after the Great Recession of 2008–2009.

2. See Bivens and Zipperer 2018 for some evidence of this.

3. See Mishel and Bivens 2021 for the central role of low-pressure labor markets in suppressing wage growth for most of the post-1979 period. See Gould 2020 for a broad overview of wage trends over the same period.

4. See Wilson 2015 for evidence on this.

5. While the differences between the coefficients are not statistically significant at most conventional levels, the difference in magnitude is economically large and is consistent and robust across the various regression specifications and time periods.

6. We switch to using an output gap measure for the state of the overall economy in this section because there is an arithmetic relationship between the overall unemployment rate and disaggregated measures of unemployment and employment by group. When, for example, the Black unemployment rate falls, this will by definition also reduce the measured overall unemployment rate. Our output gap measure does not have any arithmetic relationship to disaggregated labor force measures, so we use it for the rest of this paper.

7. See Hobbs and Stoop 2002 for evidence of this.

8. See Derenoncourt 2021 for evidence that as neighborhoods’ share of Black residents increased over time, public investments shifted more heavily toward policing and incarceration and white students saw higher enrollments in private schools. See Johnson 2011 for evidence that racial segregation led to lower resources for students in schools with higher shares of Black students.

References

Aizer, Anna, Ryan Boone, Adriane Lleras-Muney, and Jonathan Vogel. 2020. “Discrimination and Racial Disparities in Labor Market Outcomes: Evidence From WWII.” National Bureau of Economic Research (NBER) Working Paper #27689, August 2020, https://doi.org/10.3386/w27689.

Bernstein, Jared, and Janelle Jones. 2020a. “The Federal Reserve Could Help Make the Job Market Fairer for Black Workers.” Washington Post, June 15, 2020.

Bernstein, Jared, and Janelle Jones. 2020b. The Impact of the Covid-19 Recession on the Jobs and Incomes of Persons of Color. Groundwork Collaborative and Center on Budget and Policy Priorities. Policy Futures Report. June 2020.

Bivens, Josh. 2016. Why is Recovery Taking So Long—and Who’s to Blame? Economic Policy Institute. August 2016.

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

Bureau of Labor Statistics, Consumer Price Index (BLS-CPI). 2021. Public data series for various years accessed through the CPI National Databases and through series reports. Accessed January 2021.

Bureau of Labor Statistics, Current Population Survey (BLS-CPS). 2021. Public data series for various years accessed through the CPS National Databases and through series reports. Accessed January 2021.

Bureau of Labor Statistics, Productivity and Costs by Major Sector. (BLS-LPC). 2021. Public data series for various years accessed through the LPC National Databases and through series reports. Accessed January 2021.

Cajner, Tomaz, John Coglianese, and Joshua Montes. 2020. “The Long-Lived Cyclicality of the Labor Force Participation Rate.” Working Paper. October 2020.

Congressional Budget Office (CBO). 2021. Historical Data and Economic Projections (online database). Accessed February 2021.

Derenoncourt, Ellora. 2021. “Can You Move to Opportunity? Evidence from the Great Migration.” Working Paper. February 2021.

Economic Policy Institute (EPI). 2021a. Current Population Survey Extracts, Version 1.0.18, https://microdata.epi.org.

Economic Policy Institute (EPI). 2021b. State of Working America Data Library.

Flood, Sarah, Miriam King, Renae Rodgers, Steven Ruggles, and J. Robert Warren. Integrated Public Use Microdata Series, Current Population Survey: Version 7.0 [data set]. Minneapolis, MN: IPUMS, 2021. https://doi.org/10.18128/D030.V7.0

Gould, Elise. 2020. State of Working America 2019: A Story of Slow, Uneven, and Unequal Wage Growth Over the Past 40 Years. Economic Policy Institute. February 2020.

Hobbs, Frank, and Nicole Stoop. 2002. Demographic Trends in the 20^th Century. U.S. Census Bureau. November 2002.

Johnson, Rucker. 2011. Long-Run Impacts of School Desegregation and School Quality on Adult Attainments. National Bureau of Economic Research Working Paper #166664.

Mishel, Lawrence, and Josh Bivens. 2021. Identifying the Policy Levers Generating Wage Suppression and Wage Inequality. Economic Policy Institute. May 2021.

Robinson, Joan. 1936. “Disguised Unemployment.” The Economic Journal 36, no. 182): 225–237.

Shackleton, Robert. 2018. “Estimating and Projecting Potential Output Using CBO’s Forecasting Growth Model.” Congressional Budget Office Working Paper Series. February 2018.

Wilson, Valerie. 2015. The Impact of Full Employment on African American Employment and Wages. Report for the Full Employment Project at the Center on Budget and Policy Priorities. March 2015.

The Department of Labor (DOL) and the National Labor Relations Board (NLRB) enforce major worker protection laws, including the Fair Labor Standards Act, the Occupational Safety and Health Act, and the National Labor Relations Act. These statutes guarantee U.S. workers a minimum wage, a safe and healthy workplace, and the right to collective bargaining, respectively, but weak enforcement has led to pervasive and repeated violations of these laws. Despite inflation, a growing workforce, and increasingly complex workplaces, funding for agencies like the Wage and Hour Division (WHD), Occupational Safety and Health Administration (OSHA), and the NLRB has largely remained stagnant over the last decade, as shown in Figure A.

Figure A

In large part, these agencies’ inability to enforce these statutes fully can be linked to insufficient resources, which impedes the hiring of more investigators or inspectors—staff who play critical roles in initiating investigations and ensuring compliance. As shown in Figure B, the number of private-sector workers per WHD investigators and federal OSHA inspectors has largely increased since 1979. This means that each investigator or inspector is “responsible” for far more employees than they were four decades ago.

Figure B

For example, there were 61,642 workers for every federal OSHA inspector in 1979, but this number tripled to 198,532 by 2019. That is 19 times the International Labor Organization’s recommendation of one inspector per 10,000 workers. Even more concerning, OSHA had a record-low number of investigators in 2020 before the onset of the pandemic, which contributed to the sharp decline in agency investigations. OSHA conducted only 21,674 inspections in fiscal year (FY) 2020, compared with 33,393 in FY2019 and 32,023 in FY2018.

This chronic under-resourcing of enforcement agencies has been detrimental to workers’ wellbeing. In 2019, 275 workers died each day from workplace injuries, with Latinx and Black workers—who face more dangerous working conditions—having higher risks of dying on the job. Further, nearly 3.5 million workers suffered workplace injuries or illnesses, as reported by employers. While these statistics are already staggering, they likely underestimate the severity of issues workers face. In fact, the Government Accountability Office (GAO) found that the majority of employers still fail to report workplace injuries due to OSHA’s limited resources and procedures.

Unsurprisingly, the Wage and Hour Division’s enforcement has also suffered from a steady decline in critical staffing. In 1979, there were 81,717 workers for every wage and hour investigator, but that figure more than doubled to 189,878 by 2019. Further, between 2010 to 2019, the number of investigators shrunk by 25% from 1,035 to only 780, and staff interviewed by GAO directly attributed these declines to the agency’s funding levels. Consequently, officials reported taking various steps to stretch its already-thin resources and staffing, including by dropping “low-priority” cases.

Insufficient enforcement also burdens workers with remarkable monetary costs. A 2017 report estimates that in the 10 most populous states in the country, workers lose $8 billion each year to wage theft, averaging $3,300 of lost wages annually per year-round worker. In sharp contrast, DOL and state departments of labor and attorneys general in 2016 recovered just $414 million in back wages on behalf of workers. A recent report from the Peterson Institute for International Economics confirms that the DOL’s weak enforcement system does little to deter wage theft. The paper reports that since many employers are acutely aware of enforcement agencies’ stretched resources, they expect fewer investigations of labor law violations and minimal penalties if caught.

A look at funding and hiring levels at the NLRB, the agency that ensures that workers have the right to form a union, reveals similarly concerning trends: Enforcement continues to take a backseat in agencies’ priorities and has not kept up with the growth in the workforce. As shown in Figure C, despite an increase in the private-sector workforce, the number of full-time employees at the NLRB has dropped by nearly 31% from 1,789 to 1,320 between 2006 and 2019. During the same timeframe, the number of covered workers per NLRB staff increased by 50%, from one full-time employee per 74,809 workers to one full-time employee per 112,201 workers. Further, staffing levels at regional offices, which typically handle the intake of complaints filed by workers, dropped by 33% between 2010 and 2019. Relatedly, a report by the GAO released in March 2021 found a severe decline in morale across the NLRB, with only about one-third of employees agreeing that the agency has sufficient resources, while nearly half admitted to having an unreasonable workload. Unfortunately, the deliberate underfunding of the NLRB is symptomatic of widespread neglect of the importance of enforcement agencies.

Unionization efforts have also faced sustained attacks, as weak enforcement allows illegal conduct to flourish. Our research has shown that employers are charged with violating federal law in 41.5% of all union election campaigns, and one out of five union election campaigns involves a charge that a worker was illegally fired for union activity. Consequently, union elections today are characterized by employer intimidation and in no way reflect the democratic process guaranteed by the National Labor Relations Act. 

Figure C

While recent trends in enforcement funding are dismal, there is renewed political will to prioritize the just and timely enforcement of our nation’s labor laws. The Biden administration has signaled its intent to escalate investigations of workplace abuses and has put forth proposals that would significantly boost agencies’ funding. In fact, the Biden proposal sets aside at least $2.1 billion for worker protections, a significant step towards meeting the needs of today’s workforce. As we continue on the road to recovering from the pandemic, stronger enforcement of labor laws should remain central to discussions about reviving the economy and uplifting the working class.