Today’s #jobs data can be considered a solid bounce back to relative weakness in June and July. Job growth has averaged 71k over the last 3 months. Weaker numbers for leisure and hospitality and unusual July losses in local government education employment seems to have resolved in August.
#econsky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:43 AM · Sep 4, 2026

Leisure and hospitality and state/local government led the job growth for August, following by construction. Information and financial activities reported losses. Federal employment continues to trend down.
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:51 AM · Sep 4, 2026

With August losses, federal employment is now down 336,000 jobs since January 2025. The vital services federal employees provide cannot be done without these essential workers.
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:52 AM · Sep 4, 2026

The preliminary benchmark revisions were out last week, regular BLS communication needed for timely and accurate data. That release suggests there were 79,000 fewer jobs added than originally reported since Trump took office, including 178,000 fewer private sector jobs.

www.bls.gov/news.release…

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:59 AM · Sep 4, 2026

Nominal wage growth decelerated in August, rising just 3.1% over the year. Slowing nominal wage growth suggests workers don’t have the leverage to bid up their wages. Even with low unemployment, the depressed hires rate means workers aren’t finding new jobs to raise their wages.
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:15 AM · Sep 4, 2026

Even though unemployment held steady, I have continuing concerns about the depressed hires rate. Those who are lucky enough to have a job are sitting tight while new entrants or long-term employed can’t find work. Long-term unemployment has been steadily rising led by those unemployed over a year.

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:37 AM · Sep 4, 2026

Key takeaways:

• So-called right-to-work (RTW) laws undermine workers’ ability to form unions. Unionization rates are clearly lower in RTW states—and since unions raise wages, that suppression lowers pay.   

• States with RTW laws are also more likely to have other anti-worker laws and generally weaker labor standards. This is likely because both RTW and other anti-worker policies stem from the same anti-worker political roots, and because RTW’s suppression of unions deprives these states of strong potential political champions (unions) for other pro-worker policies.  

• Non-RTW states are more likely to have pro-worker policies like higher minimum wages and prevailing wage laws, restrictions on noncompete agreements, pay transparency policies, and more protective unemployment insurance benefits. 

• As a result, workers in RTW states are paid 6.7% less on average than workers in non-RTW states—more than double the 3.2% wage penalty we found in 2015. That means the RTW wage penalty has increased in recent years and now translates to over $4,000 less per year for a median full-time worker in a RTW state.

• The RTW wage penalty is steeper for women (7.3%) than for men (6.3%), and for Black (9.7%) and Hispanic workers (10.0%) than for white workers (5.4%).   

• Lawmakers already have the policy tools they need to boost wages and improve other outcomes for workers: 

• • At the state level, rolling back existing RTW laws is the most powerful lever.  

• • At the federal level, the Protecting the Right to Organize Act (which includes a ban on RTW) would be transformational for efforts to organize workers.  

• • Policymakers could also make progress on numerous other fronts, such as making UI systems more protective and raising minimum wages. 

In the United States, so-called right-to-work (RTW) laws are anti-union state policies promoted by big business interests to weaken unions and limit workers’ bargaining power. Despite the name, RTW laws do not provide any sort of job protection or right to a job. Rather, the deceptively named policy starves unions of resources, making it harder for them to negotiate collectively for better wages, benefits, and working conditions.

RTW laws emerged in the 1940s as part of anti-union campaigns to suppress worker organizing and maintain Jim Crow labor relations in Southern states; these campaigns worsened economic inequality and racial disparities (Childers 2024). Today, 26 states have active “right-to-work” statutes.1

As intended, RTW laws reduce union density (the share of workers who are either members of a union or covered by a union contract). These policies prohibit unions and employers from negotiating over union security—the contract terms under which workers covered by a union contract either join the union or pay an agency fee covering their share of representation costs. This deprives unions of critical resources and suppresses union membership (Sherer et al. 2026). RTW states—specifically those with RTW laws for at least 25 years—have an average union density rate of 6.2%, far less than half the union density rate in non-RTW states (16.0%) (see Appendix Table 1).

RTW laws not only reduce the number of workers with collective bargaining agreements but also limit unions’ political power to advocate for stronger labor standards for workers more broadly. This is, in part, why states without anti-union restrictions are more likely to pass protections such as prevailing wage mandates, restrictions on noncompete agreements, higher minimum wages, and pay transparency laws.

To assess the effects of RTW laws on outcomes for workers, we analyze whether wages are lower in RTW states after controlling for key demographic-, job-, and state-level characteristics (see Appendix Table 1 for full demographic controls as well as median and average wages by RTW status). We find that wages in “right-to-work” states are, on average, 6.7% lower than in non-RTW states. This penalty is worse for Black and Hispanic workers, as well as young workers and those without college degrees. Notably, the penalty is also larger for nonunion workers. This is because strong unions benefit workers across the state, not just those in unionized workplaces (Shierholz et al. 2026). In RTW states with weaker unions, nonunionized workers do not experience these positive “spillover” effects.

Pro-worker policies are more common in states without RTW laws

The strong correlation between wages and states’ RTW status is not surprising. Not only do RTW states have lower unionization rates, leaving workers with less leverage to bid up wages, but they often maintain other anti-worker policies. Combined, these anti-worker policies tilt bargaining power toward employers, weakening workers’ ability to garner higher wages. States without RTW laws are more likely to have a wide array of stronger labor standards in place that raise pay, protect workers’ ability to switch jobs, and support those who lose one. These same states—where unionization rates have not been suppressed by RTW laws—have a better safety net and stronger labor standards in part because stronger unions give workers more political power and the ability to better lobby for pro-worker policies writ large (Shierholz et al. 2026).

Figure A displays the incidence of these labor standards for states with and without anti-union RTW laws.2 This is far from a complete list of relevant policies that may correlate with RTW status and wage levels, but they are illustrative.

Figure A

Prevailing wage

Prevailing wage laws apply to certain contracts for public works and require that contractors and subcontractors pay the employees performing such work a prevailing wage rate. The prevailing wage rate is defined as the average wage paid to similarly employed workers in a specific occupation in the area of intended employment. Federal law requires this of federal contracts while state and local laws may apply to state and local contracts.

Prevailing wage laws are more common in non-RTW states. Figure A shows that nearly all (96.0%) of non-RTW states have prevailing wage laws, while only 23.1% of RTW states have these protections for workers on state contracts.3 Without prevailing wage requirements, contractors may reduce their workers’ wages to win bids on government contracts—putting contractors who pay their workers fair wages at a disadvantage.

Restrictions on noncompetes

Noncompete agreements are employment provisions that ban workers at a company from joining a competing business or starting their own for a set period after leaving a job. Noncompetes undermine economic dynamism and innovation; they depress business formation and labor mobility, hurt productivity and growth, raise prices, shrink workers’ wages, and restrict workers’ freedom (Shierholz 2024).

One of the most effective ways workers achieve higher wages is by securing another offer, and then either renegotiating their current salary or leaving for the new higher-paying job. It is not surprising then that workers in states that enforce noncompetes are paid less than similar workers in states that do not enforce noncompetes (Starr 2019). States without anti-worker RTW laws are nearly four times as likely to restrict noncompetes as RTW states (44.0% vs. 11.5%).4 In this analysis, we include only full bans on noncompetes and states with income-based restrictions. States with only small carve-outs for certain occupations provide a much weaker standard and are, by definition, less broad, so those aren’t included.

Minimum wage

Raising the minimum wage is particularly important for lower-wage workers. While the federal minimum wage has sat at $7.25 an hour since 2009, losing over 30% of its purchasing power, workers in 31 states enjoy a higher wage floor (Zipperer 2026). Non-RTW states are three times as likely to have minimum wages higher than the federal minimum compared with RTW states (92.0% vs. 30.8%). This is not a small difference. As of 2026, the average value of the minimum wage in non-RTW states is 60% higher—$5.36 more per hour—than in RTW states.5

Pay transparency

Pay transparency laws prevent employers from lowballing wage offers by requiring them to include wage information in job postings. While they vary, all laws include some requirement that employers provide salary information in job postings or if directly requested by applicants. Lack of knowledge about wages and benefits keeps potential workers in the dark and limits their ability to learn about and apply for better paid opportunities. A study on Colorado’s newly passed wage transparency law found that wages increased 4.2% faster than those in neighboring states without such laws (Shedge 2025). Further, wage transparency has the potential to reduce gender- and race-based discrimination by arming jobseekers with more information and limiting employers’ ability to pay different amounts to similarly qualified candidates. Only two RTW states have a pay transparency law, compared with 64% of non-RTW states.6

Unemployment insurance

Access to unemployment insurance (UI) also shapes workers’ ability to seek better wages. When a worker is laid off and their household income falls, unemployment insurance supports them and their family until they find another job. UI is a joint federal-state program that relies on state UI systems to effectively deliver benefits to unemployed workers. States differ in their rules and approaches under the federal UI framework, leading to wide variation in the share of unemployed workers receiving UI benefits (the UI recipiency rate).

Better UI not only cushions workers and their families but also provides workers with additional leverage in the labor market. When UI access and generosity fall, job seekers are forced to settle for lower wages and employers leverage this by offering less (Dahl and Knepper 2026). On the flip side, research shows a positive relationship between more UI generosity and wages (Rinz and Wasser 2026). The UI recipiency rate is significantly higher in non-RTW states compared with RTW states (32.8% vs. 18.7%).7 This is not surprising because non-RTW states have stronger unions, which fight to increase UI eligibility, ease of access, benefit levels, and benefit duration—to the benefit of all workers, union and nonunion (Hertel-Fernandez and Gould-Werth 2020). As with other pro-worker policies, better UI access can contribute to higher wages in non-RTW states.

Other policy levers

Not only do workers in non-RTW states enjoy higher wages, but they also benefit from increased economic security thanks to higher health insurance rates, higher public education spending, paid family and medical leave access, and no preemptions against higher labor standards. Every non-RTW state expanded Medicaid compared with fewer than two-thirds of RTW states (61.5%).8 Unsurprisingly, the uninsured rate is higher in RTW states than in non-RTW states (8.9% vs. 6.2%).9

Investments in public education deliver a more productive workforce and more informed and engaged society. Increased funding for schools not only improves educational attainment, but also increases wages and family incomes (Jackson, Johnson, and Persico 2016). Non-RTW states spend 45% more on public education than RTW states ($24,084 vs. $16,620).10 Paid family and medical leave provides essential benefits that help workers maintain their livelihoods while taking care of themselves and their families. These laws are also shown to increase labor force participation, job retention, and earnings (Glynn 2020). Only one RTW state has comprehensive paid family and medical leave provisions for private-sector workers, compared with 56% of non-RTW states.11

Finally, some cities, counties, and other local governments enact policies that raise standards for working people, but some state legislatures force those standards back down with preemption—the use of state law to void local ordinances. Preemptions apply to policies such as minimum wage, prevailing wages, paid leave, and fair scheduling (EPI 2025). Preemption laws interfere with local governments’ ability to set job quality standards. They suppress wages, exacerbate racial inequities, and reduce worker power (Sherer, Cohn, and Ahdoot 2025). These preemptions are more common in RTW states.

Workers in RTW states are paid less

Our analysis of the relationship between wages and RTW status puts states into three categories: long-term RTW states, always non-RTW states, and “switcher” states that have changed their RTW status since 2011.12 Appendix Table 1 provides descriptive statistics for these three groups of states, and Appendix Figure A provides a map of states by RTW status. Our estimates focus on comparisons between long-term RTW states and always non-RTW states.

A simple comparison of hourly wages, the primary variable of interest, reveals that average wages are 22.9% higher in non-RTW states ($40.28 in non-RTW states vs. $32.78 in RTW states). Median wages are 19.2% higher in non-RTW states ($28.79 vs. $24.16). Because there are differences between worker-, job-, and state-level characteristics in RTW and non-RTW states, and since some of these characteristics will directly impact workers’ wages, it is important to control for these factors in a multivariate regression model. This allows us to more accurately identify the relationship between RTW status and wages.

We estimate log wage equations using Bureau of Labor Statistics Current Population Survey Outgoing Rotation Group (CPS-ORG) data for 2023–2025. We pool three years of data to minimize any spurious year-specific economic relationships, thereby helping us achieve more precise estimates. The total sample consists of 234,155 workers, ages 16+, who earn wages and salaries. About 42% of the sample lives in states with RTW laws (see Appendix Table 1 for the sample’s full demographic breakdown).13

In Table 1, we construct a regression model, starting with an unadjusted framework and building up to a model that controls for the full range of explanatory variables. The dependent variable is the natural log of hourly wages.

The variable of interest is an indicator variable when the worker lives in a long-term RTW state. We also include an indicator in the regression for a switcher state, though that is not our focus of interest (see Appendix Table 2 for a complete set of regression results, including the coefficient for switcher states). There are currently 26 states with RTW laws on the books and 25 without them including Washington, D.C. (see Appendix Figure A). There are five switcher states that have adopted RTW in recent years: Indiana (2012), Michigan (2013), Wisconsin (2015), West Virginia (2016), and Kentucky (2017). In 2023, Michigan became the first state to repeal its RTW statute. We separate the five switcher states from the analysis to better isolate the relationship between wages and the long-run effects of RTW.

The naïve model, without any controls, suggests that workers in RTW states are paid 17.8% less than workers in non-RTW states (Model I in Table 1).14 But many factors contribute to wage differences: demographic characteristics such as age, race/ethnicity, gender, and level of education are relevant as well as job-related factors such as work hours, occupation, and industry. Since these vary across states, they reduce the measured relationship between RTW status and wages. Model II shows that when the analysis compares more similar workers with each other, the RTW penalty drops to 14.0%.15

There are other factors that are different in RTW states, aside from these individual and job-related characteristics. Most importantly, the cost of living. The cost of living in non-RTW states is higher than in RTW states, as shown in Appendix Table 1. In prior research, Gould and Kimball (2015) showed that choice of price indicator makes no material difference on the measured relationship between RTW and wages. Here we include the Bureau of Economic Analysis’ Regional Price Parity index: The regional price parity index is higher in non-RTW states.16 Not surprisingly, this correlates with wage levels and therefore is important to include in the model. As with earlier studies, we also include state unemployment rates to control for differences in economic conditions.

After including the full set of demographic, individual-, and state-level labor market and cost of living controls (Model III), our analysis finds that workers in RTW states are paid 6.7% less than similar workers in non-RTW states. For the average worker, this is a difference of $2.70 per hour, or more than $5,600 for full-time workers at 2,080 hours over the year. However, because the average is skewed upward due to wage inequality, an assessment at the middle of the wage distribution provides a more conservative but perhaps more robust estimate of the RTW wage penalty. Acknowledging that this is a likely understatement of the full relationship, the RTW wage penalty for the median worker at full-time hours is over $4,000 for the year.

Table 1

This does not mean that RTW laws are a causal explanation for this 6.7% wage penalty. As demonstrated earlier in this report, RTW status correlates strongly with a number of other anti-worker policies. Other research has attempted to calculate the causal effects of RTW laws on wages. Using the switcher states in the most recent period, Fortin, Lemieux, and Lloyd (2022) found smaller wage penalties, which is to be expected given the shorter time frame for measurement post-adoption. Our wage penalty for switcher states—shown in the final column of Appendix Table 2—is much smaller (closer to zero) than the coefficient for long-term RTW states, though not statistically significant at conventional levels. Dasgupta and Merchant (2023) also examine states recently adopting RTW laws and find a statistically significant decline in annual wages by almost $1,900. Again, these states are more recent RTW adopters, so the full effects we find may take time to materialize.

We find that this relationship between long-term RTW status and wages remains economically and statistically significant under alternative specifications of our econometric model. For instance, removing major industries and occupations changes the coefficient to -0.071. Restricting the model to RTW status as of 2026—removing the separate indicator for states that switched since 2011—yields a coefficient estimate of -0.060. Doing the same but dropping Michigan (since its status changed in the measurement period) results in a coefficient of -0.064. In our final model, we do not include an indicator for unionization, unlike Gould and Kimball (2015). If we reintroduce that variable, the coefficient is -0.063.

In each of our regressions discussed above, we include a three-year average as it improves data reliability and allows us to run smaller groups of the data (see next section). When we run our analysis using just 2025, our results do not change. In fact, when we run our three-year final model for the periods 2021–2023, 2022–2024, and 2023–2025, our coefficients remain in the -0.070 to -0.068 range.

Wage penalties are steeper for women, Black, Hispanic, and nonunion workers in RTW states

The RTW penalty is not uniform across demographic groups and other labor market characteristics. Table 2 below shows the results of a series of final-model regressions (comparable to Model III in Table 1) for a set of demographic groups.

We find that women’s wages in RTW states are penalized at a higher rate (7.3%) than men’s (6.3%). The wage penalty also persists across all racial and ethnic groups, but it is almost twice as large for Black (9.7%) and Hispanic (10.0%) workers as for similar white workers (5.4%) and more than twice as large as for Asian American and Pacific Islander (AAPI) workers (4.6%). Young workers between the ages of 16 and 24 face a greater penalty (8.4%) than older workers. And workers without a college degree experience a much higher penalty (7.8%) than workers with a college degree (4.7%).

Table 2

It is particularly important to note that RTW laws have a statistically significant negative effect on the wages of nonunion workers, as shown in the last row of Table 2. Our analysis indicates that nonunion workers in RTW states have wages that are 6.3% lower, on average, than their counterparts in non-RTW states. Union members likely experience a smaller penalty (3.6%) because their wages tend to be higher than those in nonunionized workplaces, regardless of the state’s RTW status (McNicholas et al. 2025). However, the wage impact of unions extends beyond those directly covered by a union contract. Unions lift wage standards across entire industries, meaning that nonunion workers in states with higher union density are more likely to receive better pay.

As the gap in wages and labor standards between RTW and non-RTW states has grown, so has the wage penalty

Using our final model (Model III in Table 1) with pooled 2023–2025 data, we find a 6.7% RTW wage penalty—a larger penalty than we found in earlier examinations. For example, we found a 3.2% wage penalty using data for 2009 (Gould and Shierholz 2011) and pooled 2010–2012 data (Gould and Kimball 2015). Some of the increase in the current estimate of the RTW wage penalty reflects small changes (likely improvements) in our methods. But part of the higher estimate persists even when we apply our current methods to earlier data: Applying these methods to the 2010–2012 data yields an RTW wage penalty of 4.7%.

One method change concerns “switchers”—states that changed from non-RTW to RTW status (or back) since 2019. In the 2010–2012 period, there were no switchers to affect the data.17 In our analysis, we isolate switchers from the long-term RTW group, but the RTW penalty remains negative and significant no matter how one deals with the issue of switcher states. For example, recategorizing the five switchers to match their 2012 status (all non-RTW) lowers the measured RTW wage penalty in the 2010–2012 period to 4.1%, but it remains significant.

Other changes relative to our earlier research include controlling for public-sector status; removing union as an independent variable from the regression; using the full working age range of 16 and up (as opposed to 18–64); and correcting an issue with nonrandom missing values to the metropolitan variable. We’ve further clustered standard errors at the state level as suggested by Douglas (2024), which does not affect the size of the coefficients, only the size of their standard errors and resulting statistical significance.18

Setting those minor model differences aside, a consistent comparison over time sees the RTW wage penalty rising from 4.7% in 2010–2012 to 6.7% in 2023–2025. One possible reason why this RTW wage penalty grew could be the increasingly divergent policy trajectories of RTW and non-RTW states over this period.

For example, the gap between minimum wages in RTW states and non-RTW states has grown since 2010, largely because policymakers in most RTW states have refused to increase their minimum wage despite the fact that the federal minimum wage has been stagnant in nominal terms since 2009. In that year, the average minimum wage in RTW states was $7.25 while the average minimum wage in non-RTW states was $7.54. That gap has grown enormously, from just $0.29 per hour in 2009 to $5.36 per hour by 2026 ($14.24 in non-RTW states vs. $8.89 in RTW states).

There were other potentially relevant policy changes between 2012 and 2025. The first state-level pay transparency law was passed in Maryland in 2020. Seventeen states followed suit in subsequent years, 15 of which were non-RTW states. On the flip side, six states have repealed their prevailing wage laws since 2012—all of which were RTW. Only one state, Michigan, reinstated its prevailing wage law. It did so in 2023, the same year it repealed RTW. While this is far from an exhaustive list, it’s not surprising that the measured RTW wage penalty has grown over time.

Repealing anti-union “right-to-work” laws and strengthening labor law is key to raising wages

The wage penalty for workers in RTW states has grown since the last time we conducted this research a decade ago. Long-standing and more recently enacted anti-union RTW laws in 26 states, in tandem with other anti-worker policies, have further eroded workers’ collective bargaining power, suppressed wages, and lowered labor standards. These outcomes are linked to clear policy choices, and lawmakers at every level of government have the tools they need to restore workers’ union rights, strengthen labor standards, and lift wages. Across the U.S., 56 million workers say they want a union in their workplace, yet only 10% of all workers have one (Shierholz et al. 2026). Current federal and state labor laws—including anti-union RTW laws—place too many obstacles in workers’ path to unionizing and give employers too much power to interfere with workers’ free choice.

First and foremost, states with RTW laws should follow Michigan’s recent lead and repeal them. Repealing state RTW laws will enable workers to unionize and raise wages across the country. Indeed, recent EPI research finds that equalizing collective bargaining rights across all states by repealing RTW laws that affect private employees and ensuring collective bargaining rights for public employees would increase national union density by almost 50% (from 9.9 to 14.4%) (Shierholz et al. 2026). States should also consider passing constitutional amendments that affirm collective bargaining rights and explicitly bar RTW-style restrictions, such as the 2022 Workers’ Rights Amendment in Illinois (Sherer 2026).

At the federal level, lawmakers should pass the Protecting the Right to Organize (PRO) Act. The PRO Act is designed to address major weaknesses of the National Labor Relations Act (NLRA), which has been severely eroded by amendments and court decisions. Among many other reforms, the PRO Act would eliminate the option for states to maintain anti-union RTW laws—restoring full bargaining rights to workers in all states as intended when the NLRA was originally passed in 1935 (McNicholas, Poydock, and Rhinehart 2021). Since its first introduction in 2019, the PRO Act has passed the House of Representatives twice with bipartisan support (Shierholz et al. 2026).

In addition to proposed PRO Act reforms that would enable newly unionized workers to pursue arbitration to achieve a first contract (if an employer fails to negotiate in good faith), federal labor law should set a minimum standard that such contracts include a cost-of-living adjustment (COLA). To further expand the benefits of collective bargaining, Congress should amend the NLRA to require default collective bargaining at any firm where CEO-to-worker pay ratios exceed 100 times what a typical worker makes in their industry (as this ratio would be one proxy for whether conditions at an employer are particularly unequal) (Shierholz et al. 2026). 

Legislators at both the state and federal level should also look to the list of anti-worker laws that are typically found in RTW states for more opportunities to rebalance unequal power in the labor market and improve wages and working conditions. These policies include:

• Raising the minimum wage;
• Banning noncompete agreements;
• Strengthening federal unemployment insurance guidelines and state UI systems;
• Removing state restrictions that preempt local governments from raising wages and strengthening labor standards.

These are among just a few of the policy levers states can use to raise wages and standards (EPI 2026b).

Our analysis suggests that repealing anti-union RTW laws and enabling workers to increase unionization levels are key to raising wages, both because RTW laws directly suppress unionization rates (and unions raise wages), and because weakening unions limits workers’ collective ability to shape other public policies.

Acknowledgements

The authors are grateful to Josh Bivens, Jennifer Sherer, Hilary Wething, and Ben Zipperer for their helpful comments and expertise.

Appendix

Appendix Figure A

Appendix Table 1

Appendix Table 2

Notes

1. See Appendix Figure A for RTW status. As of 2026, these states are Alabama, Arizona, Arkansas, Florida, Georgia, Iowa, Idaho, Indiana, Kansas, Kentucky, Louisiana, Mississippi, North Carolina, North Dakota, Nebraska, Nevada, Oklahoma, South Carolina, South Dakota, Tennessee, Texas, Utah, Virginia, Wisconsin, West Virginia, and Wyoming.

2. States are categorized as “RTW” or “non-RTW” based on their status as of 2026.

3. Authors’ analysis of prevailing wage laws from DOL (2023) and state legislative websites.

4. Authors’ analysis of noncompete restriction policies from EIG (2026).

5. Authors’ analysis of minimum wage data from EPI (2026c).

6. Authors’ analysis of pay transparency laws from GovDocs (2026).

7. Authors’ analysis of UI recipiency rate data from DOL-ETA.

8. Authors’ analysis of Medicaid expansion policies from KFF (2026).

9. Authors’ analysis of uninsurance rates by state, 2023 through 2025, from Carter (2025).

10. Authors’ analysis of per-pupil education spending data from NCES-NPEFS (2024).

11. Authors’ analysis of paid family and medical leave policies from Wielk (2026).

12. Prior to 2012, the last state to change its status was Oklahoma in 2001.

13. This analysis builds on earlier work by Gould and Kimball (2015); see the report for a full description of the methodology and list of controls.

14. Interpreting the results from these semilogarithmic functions requires utilizing the exponential function on the coefficient. Specific to the binary variable coefficient (β1) for RTW, the percent change in workers’ wages resulting from a state being RTW can be calculated by the formula: 100*[exp(β1)-1]. Typically, the result of this equation will be very close to the coefficient itself but will differ more as the coefficient becomes larger. We show the coefficient in the first row of Table 1 and the exponentiated result as a percent in the third row.

15. Full regression results for covariates in all three models can be found in Appendix Table 2.

16. Because the 2025 BEA RPP data were not yet available at the time of publication, we assign an average of 2022–2024 data to all years in our sample (2023–2025).

17. Gould and Kimball (2015) classify Indiana as non-RTW as its RTW law had just taken effect.

18. We do not control for state-level fixed effects, as Douglas (2024) suggested, because state-level fixed effects perfectly correlate with RTW status and would deem our results uninterpretable.

References

Bureau of Economic Analysis, Regional Price Parities by State and Metro Area: All items (BEA-RPP). Various years [Release date: February 19, 2026]. Public data series accessed via the Regional Data interactive tool. Accessed August 1, 2026. 

Carter, Caitlin. 2025. “Health Insurance Coverage by State: 2023 and 2024.” U.S. Census Bureau, September 11, 2025. 

Childers, Chandra. 2024. The Evolution of the Southern Economic Development Strategy: Rooted in Racism and Economic Exploitation: Part One. Economic Policy Institute, May 2024. 

Dahl, B. Gordon, and Matthew Knepper. 2026. “Unemployment Insurance, Starting Salaries, and Jobs: Evidence from Multi-state Firms.” The Review of Economic Studies, rdag056. https://doi.org/10.1093/restud/rdag056.

Dasgupta, Kabir, and Zofsha Merchant. 2023. “Understanding Workers’ Financial Wellbeing in States with Right-to-Work Laws.” FEDS Notes (The Federal Reserve), September 8, 2023.

Department of Labor (DOL). 2023. “Dollar Threshold Amount for Contract Coverage Under State Prevailing Wage Laws.” Last modified January 1, 2023. 

Department of Labor, Employment and Training Administration (DOL-ETA). Various years. Unemployment Insurance Chartbook. Public data series accessed via the interactive tool. Accessed August 6, 2026. 

Douglas, Christopher C. 2024. Right-to-Work States Do Not Have Lower Wages. Mackinac Center, December 2024. 

Economic Innovation Group (EIG). 2026. “State Noncompete Law Tracker.” Last modified March 24, 2026. 

Economic Policy Institute (EPI). 2025. “Workers’ Rights Preemption in the U.S.” Last modified February 2025. 

Economic Policy Institute (EPI). 2026a. Current Population Survey Extracts, Version 2026.7.8, https://microdata.epi.org.

Economic Policy Institute (EPI). 2026b. Holding the Line: State solutions to the U.S. Worker Rights Crisis.

Economic Policy Institute (EPI). 2026c. “Minimum Wage Tracker.” Last Modified July 1, 2026. 

Fortin, Nicole, Thomas Lemieux, and Neil Lloyd. 2022. “Right-to-Work Laws, Unionization, and Wage Setting.” National Bureau of Economic Research Working Paper no. 30098, June 2022. http://www.nber.org/papers/w30098. 

Glynn, Sarah Jane. 2020. “The Rising Cost of Inaction on Work-Family Policies.” Center for American Progress, January 21, 2020.

Gould, Elise, and Will Kimball. 2015. “Right-to-Work” States Still Have Lower Wages. Economic Policy Institute, April 2015.

Gould, Elise, and Heidi Shierholz. 2011. The Compensation Penalty of “Right-to-Work” Laws. Economic Policy Institute, February 2011. 

GovDocs. 2026. “Pay Transparency Laws by State and Province.” Last modified July 2026.

Hertel-Fernandez, Alexander, and Alix Gould-Werth. 2020. Labor Organizations and Unemployment Insurance: A Virtuous Circle Supporting U.S. Workers’ Voices and Reducing Disparities in Benefits. Washington Center for Equitable Growth, October 2020.

Jackson, C. Kirabo, Rucker C. Johnson, and Claudia Persico. 2016. “The Effects of School Spending on Educational and Economic Outcomes: Evidence from School Finance Reforms.” The Quarterly Journal of Economics 131, no. 1: 157–218. https://doi.org/10.1093/qje/qjv036.

KFF. 2026. “Status of State Medicaid Expansion Decisions.” Last modified May 21, 2026. 

McNicholas, Celine, Margaret Poydock, and Lynn Rhinehart. 2021. “How the PRO Act restores workers’ right to unionize” (fact sheet). Economic Policy Institute, February 4, 2021. 

McNicholas, Celine, Margaret Poydock, Heidi Shierholz, and Hilary Wething. 2025. Unions Aren’t Just Good for Workers—They Also Benefit Communities and Democracy. Economic Policy Institute, August 2025.

National Center for Education Statistics, National Public Education Financial Survey (NCES-NPEFS). 2024. Public data series accessed via the Common Core of Data database. Accessed July 1, 2026. 

Rinz, Kevin, and David Wasser. 2026. “Unemployment Insurance Generosity and Wage Determination.” Federal Reserve Bank of Cleveland Working Paper no. 26-11, May 2026. https://doi.org/10.26509/frbc-wp-202611.

Shedge, Nirav. 2025. Pay Transparency Laws and Wage Dynamics: Evidence from Colorado. Social Science Research Network, November 2025. 

Sherer, Jennifer. 2026. Rights to Unionize and Collectively Bargain: State Solutions to the U.S. Workers Rights Crisis. Economic Policy Institute, February 2026.

Sherer, Jennifer, Emma Cohn, and Ruby Ahdoot. 2025. “Updated EPI Tracker Shows More States Obstructing Progress on Workers’ Rights: Harmful Preemption Laws are Increasing Inequality and Repressing Democracy.” Working Economics Blog (Economic Policy Institute), March 6, 2025.

Sherer, Jennifer, Nina Mast, Elise Gould, and Emma Cohn. 2026. Everything You Need to Know About Anti-Union “Right-to-Work” Laws. Economic Policy Institute, forthcoming 2026.

Shierholz, Heidi. 2024. “Banning Noncompete Agreements.” Testimony before the U.S. Senate Banking, Housing, and Urban Affairs Subcommittee on Economic Policy, July 30, 2024. 

Shierholz, Heidi, Celine McNicholas, Josh Bivens, Jennifer Sherer, Ben Zipperer, and Margaret Poydock. 2026. The Case for Tripling Union Membership: How Rebuilding Union Power Would Strengthen Workers, the Economy, and Our Democracy. Economic Policy Institute, July 2026. 

Starr, Evan. 2019. “Consider This: Training, Wages and the Enforceability of Covenants Not to Compete.” ILR Review 72, no. 4 (August): 783–817. https://doi.org/10.1177/0019793919826060.

Wielk, Emily. 2026. “State Paid Family Leave Laws Across the U.S.” Bipartisan Policy Center. Last updated April 23, 2026. 

Zipperer, Ben. 2026. Setting High Standards for a Federal Minimum Wage. Economic Policy Institute, May 2026. 

Weak #jobsreport all around
-Nonfarm payrolls fell by 23,000 (mostly driven by state and local losses)
-The unemployment rate ticked down for the “wrong” reasons as labor force participation and employment-to-population ratio softened
-Nominal wage growth decelerated to 3.2% over the year
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:47 AM · Aug 7, 2026

Large losses in the public sector in July (-53k), particularly in local education (-49.6k jobs). An initial look doesn’t suggest an issue with seasonal adjustment because losses were also registered in NSA data for June and July. Local education jobs have fallen by nearly 100k (-98.3k) since March.

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:01 AM · Aug 7, 2026

While public sector weakness in July was centered around state and local jobs, federal jobs ticked down by 3,000. Federal employment remains 327,000 below its January level. The vital services federal employees provide cannot be done without these essential workers (e.g. food inspectors).
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:12 AM · Aug 7, 2026

Although manufacturing employment ticked up slightly in July (+5k), manufacturing jobs are still down by 62k since January 2025.

Instead of adding these blue collar jobs, the manufacturing sector has lost 62,000 jobs since Trump took office.

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:16 AM · Aug 7, 2026

Nominal wage growth decelerates in July, rising just 3.2% over the year. Slowing nominal wage growth suggests workers don’t have the leverage to bid up their wages. And, along with rising prices, workers and their families continue to find it difficult to make ends meet.

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:26 AM · Aug 7, 2026

Today's jobs report came in weaker than expected. The economy added 57k jobs in June and prior months were revised down. April and May are a combined 74k lower than previously reported. Analysts believe the World Cup added 40k to June numbers—without that, job growth would have been 17k.
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) July 2, 2026 at 8:48 AM

The household survey came in even weaker. While the topline unemployment rate ticked down, it happened for the wrong reasons as labor force participation fell by 720k while employment fell by 507k. Even the prime-age employment-to-population ratio, which had remained resilient, fell 0.6ppts in June.

[image or embed]

— Elise Gould (@elisegould.bsky.social) July 2, 2026 at 8:58 AM

Given expectations around the World Cup, it's surprising to me that leisure and hospitality fell by 61k in June (and May growth was revised down by 30k). Perhaps those gains are offset by reduced discretionary spending as real wages fall.
#NumbersDay #EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) July 2, 2026 at 9:18 AM

We won't get the inflation data for June until July 14, but recent price data suggest year over year real wages likely fell in June. Workers and their families are finding it increasingly difficult to make ends meet and real wages are most surely now below where they were in January 2025.
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) July 2, 2026 at 9:18 AM

Manufacturing employment is crawling along, gaining 3k jobs in June, all in durable goods. After downward revisions, manufacturing lost jobs in May.

Since January 2025 when Trump took office, the manufacturing sector has lost 75,000 jobs.

#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) July 2, 2026 at 9:30 AM

While there's been little change this year, federal employment has shrunk an alarming 324,000 jobs since January 2025. The vital services federal employees provide cannot be done without these essential workers.
#NumbersDay #EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) July 2, 2026 at 9:31 AM

The latest jobs report came in stronger than expected this morning. The economy added 172,000 jobs in May and the unemployment rate held steady at 4.3%. Nominal wage growth continued to decelerate, further exacerbating affordability as prices rise.
#EconSky #NumbersDay @epi.org

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:46 AM · Jun 5, 2026

Job growth was strongest in leisure and hospitality, state and local governments, and health care. Losses continued in financial activities in May.

[image or embed]

— Elise Gould (@elisegould.bsky.social) 7:58 AM · Jun 5, 2026

Manufacturing employment rose by 7,000 in May, slowly clawing back the large losses last year.

Since January 2025 when Trump took office, the manufacturing sector has lost 68,000 jobs.

#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:00 AM · Jun 5, 2026

While there’s been little change this year, federal employment has shrunk an alarming 333k jobs since Jan 2025. The vital services federal employees provide cannot be done without these essential workers.
#NumbersDay #EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:08 AM · Jun 5, 2026

Nominal wage growth continued to slow in May, now 3.4% over the year. While we don’t get the May inflation data until next week, it’s very likely, given recent trends, that real wages will continue to fall and workers and their families will find it increasingly difficult to make ends meet.
#EconSky

[image or embed]

— Elise Gould (@elisegould.bsky.social) 8:14 AM · Jun 5, 2026

Key takeaways:

• Misclassification of workers as independent contractors is a pervasive and widespread problem. AAPI workers are overrepresented in three of the 11 commonly misclassified occupations: manicurists and pedicurists, home health aides, and personal care aides. Vietnamese, Bangladeshi, Filipino, Samoan, and other Pacific Islander workers are overrepresented within these occupations.
• Groups with lower median hourly wages also have larger shares of their working populations in the 11 commonly misclassified occupations.
• Federal protections against misclassification are limited and currently under attack by the Trump administration. The state and local landscape for curbing misclassification is varied, which leaves some workers less protected than others.

In March, EPI published updated research highlighting the cost to workers of being misclassified as an independent contractor for 11 commonly misclassified occupations. Asian American and Pacific Islander (AAPI) workers were overrepresented in three of those occupations—manicurists and pedicurists, home health aides, and personal care aides—relative to their share of the overall workforce.

Most federal, state, and local labor laws apply only to employees and not to independent contractors, so misclassification strips workers of key protections such as minimum wage laws or qualifying for employer-provided health insurance and retirement benefits. Additionally, both misclassified workers and social insurance funds lose out on income: the report conservatively estimates that for the three jobs in which AAPI workers are overrepresented, misclassification costs workers at least $7,000 annually and costs social insurance programs $600 to $800 per worker each year.

With the understanding that the umbrella term “AAPI” encompasses an immensely diverse population both in ethnic origin but also in economic outcomes, this piece goes beyond the narrow view that all AAPI workers are high-wage earners. Below, we provide more detail on which groups of AAPI workers are most likely to be employed in lower-wage commonly misclassified occupations.

Disaggregated data shed light on particular AAPI communities that may be vulnerable to misclassification

Across all occupations, AAPI workers comprise approximately 8% of the total workforce. For three of the 11 occupations highlighted in the report—manicurists and pedicurists, home health aides, and personal care aides—AAPI workers make up 67%, 13%, and 10% of employment, respectively, according to Current Population Survey (CPS) data.

Table 1 provides a detailed breakdown of the composition of the AAPI workforce for the three occupations in which AAPI workers are overrepresented. Here, we use the American Community Survey (ACS) as it offers detailed race definitions which the CPS does not offer due to sample size restrictions.

Asian Indian and Chinese populations combined make up over 40% of the working-age AAPI population, thus their relatively large shares of the AAPI workforce in these occupations are not surprising. However, several groups are disproportionately represented across these occupations compared with their share of the overall AAPI workforce.

For example, Bangladeshi workers make up 5.1% of AAPI workers employed as home health aides while only constituting 1.1% of the total AAPI workforce. Chinese workers represent almost half (47.7%) of AAPI home health aides while representing just over one-fifth of the overall AAPI workforce (20.9%). AAPI employment among manicurists and pedicurists is largely held by those of Vietnamese origin (71.4%).

Finally, a majority of AAPI personal care aides are either Filipino (32.8%) or Chinese (20.8%). Filipino workers, however, are overrepresented by twice their share of the overall workforce. While Samoans and other Pacific Islanders comprised a much smaller share of personal care aide employment, they are also overrepresented in this occupation by more than twice their share of the overall workforce.

Table 1

Figure A provides a more comprehensive picture of the share of each detailed group employed across all 11 commonly misclassified occupations, revealing that smaller communities—often overlooked because of their size relative to the aggregate AAPI workforce—may be among the most vulnerable to misclassification. Workers belonging to seven of those groups are more likely than the average U.S. worker to be employed in one of those occupations. Almost 20% of Vietnamese workers are employed in one of those occupations, with over half concentrated as manicurists and pedicurists.

Samoan, Hawaiian, and other Pacific Islanders have the next highest shares working in the 11 occupations, making up 15% or more of their total working-age population. These groups also earn lower median hourly wages than the national median and the aggregate AAPI median hourly wage. Their disproportionate representation in commonly misclassified occupations further exposes these workers to wage suppression due to misclassification.

Misclassification enforcement varies by state—meaning different AAPI populations can be disproportionately impacted

Federal protections from misclassification are limited and are currently under attack by the Trump administration, which has proposed a rule to weaken standards to determine worker classification under the Fair Labor Standards Act, the Family and Medical Leave Act, and the Migrant and Seasonal Agricultural Protection Act. The proposed rule narrows the definition of who is a covered employee under these statutes, encouraging employer schemes to reclassify their employees as independent contractors to evade those obligations.

Broadly, the Trump administration has been actively dismantling long-standing federal worker protections, leaving states to bear the responsibility of ensuring workers are given rights and protections and that they can exercise them. For most states, labor and employment protections only apply to workers classified as employees, meaning workers misclassified as independent contractors are denied their legal rights and protections.

EPI’s 2026 misclassification report outlines state and federal policy recommendations that ensure proper enforcement mechanisms to curb misclassification. One of the recommendations includes implementing the ABC test. Unlike the six-part “economic reality” test or the “common law” test, the ABC test presumes that a worker is an employee unless they can demonstrate they are an independent contractor based on three criteria. Placing the onus on the employer to determine the employment status of a worker provides protections against misclassification and extends proper protections to workers. Many states have adopted the ABC test for unemployment insurance programs and, to a lesser extent, for wage and hour orders and other employment applications.

As shown in Figure B, The AAPI population is highly concentrated across a handful of states. Almost half of the prime-age working Asian population is concentrated in California, New York, and Texas, and a majority of the Pacific Islander population resides in California, Hawaii, and Washington. Overall, 21 states have significant numbers of AAPI residents, and some are home to large shares of specific AAPI communities. For example, the Hmong community in Minnesota and the Burmese community in Indiana are concentrated in states that have smaller total AAPI populations.

The current landscape for state policy protections against misclassification is quite varied. For example, among the states with the largest AAPI populations, California is the only state to adopt the ABC test for both unemployment insurance and employment law, although certain occupations are exempt from the test—including app-based drivers. California also institutes penalties for misclassifying a worker, which can include restitution payments and, if the misclassification was willful, a penalty between $5,000 to $25,000 per violation.

Texas, on the other hand, has significantly less state enforcement. Apart from using the common law test for its unemployment insurance program and providing a definition of an independent contractor for workers’ compensation, Texas mainly relies on federal law for classifying workers as employees. In the last 15 years, Texas lawmakers have introduced several bills that would create penalties for misclassifying workers in the construction industry, but all have stalled or failed.

Figure B

Comprehensive protections are needed to protect workers from misclassification

AAPI workers are facing multi-pronged attacks from the Trump administration through the degradation of federal protections for workers, immigration, and equity. Occupational segregation and other labor market disparities lead women, people of color, and immigrants to be disproportionately represented in occupations that are commonly misclassified. These factors—in addition to historical and current geopolitical relations that shape the flow of labor to the U.S., immigration and citizenship status, and English language proficiency—can contribute to the concentration of AAPI workers in these occupations. Disaggregated data further identify which specific AAPI communities are overrepresented, revealing that smaller, less economically secure groups are often most exposed to the costs of misclassification. Strong policies at the federal, state, and local levels are needed to combat misclassification and to ensure workers can exercise their rights.

Key takeaways:

• The vast majority (85%) of young college graduates work in occupations that have seen strong employment growth in recent years.

• Young college graduates, like college graduates in general, are more likely to work in AI-exposed occupations than the overall workforce—and considerably more likely than young noncollege workers.
• But both young college graduates and young noncollege workers have experienced rising unemployment over the last three years, suggesting AI is not likely to be driving labor market weakness.

In the first blog post of our Class of 2026 series, we showed that the strong labor market for young college graduates of the early 2020s had begun softening in recent years. A growing share of young college graduates are seeking employment, but because their employment rates have not kept up with this job search, their unemployment rate has risen faster than the overall rate. The second blog post in the series discussed the industries where young college graduates worked. We found that recent graduates work in growing industries, but are forced to enter a weakened labor market with less job turnover, deteriorating their ability to break in. Young college graduates work in the tech sector at a similar rate to college graduates, and there is no clear evidence that tech sector employment is significantly decreased despite warnings about the advancement of AI.

In this blog post, we delve deeper into the occupations where young college graduates are likely to work.1 We examine whether it has been relatively more difficult to secure employment in these fields as the labor market has weakened. We also scour the data for signs that exposure to AI-related occupations may disproportionately affect the prospects for young college graduates as they enter the labor market.

Most young college graduates work in occupations with strong growth

Over 60% of young college graduates work in professional and related occupations or management, business, and financial occupations. Figure A displays the share of employment in each occupation or occupation grouping for young college graduates ages 22 to 27, all college graduates, and young workers without a four-year college degree. Occupations in the figure appear in order of the share of young college graduates employed in each, from largest to smallest. Over half (62.8%) of young college graduates work in professional, management, business, and financial occupations. Workers of any age with a college degree are slightly more likely to work in those two occupations (64.5%), though more likely in management occupations than professional occupations. On the other hand, nearly half (48.3%) of young noncollege workers are in service occupations or farming, construction, installation, and production occupations.

Figure A

Figure B shows the change in employment in each occupation between 2019 and 2026 and between 2023 and 2026, arranged in the same order as Figure A for comparison. Since 2019, management, business, and financial occupations and transportation and material moving occupations experienced the most growth, followed by professional and related occupations.

The top four occupations for job growth since 2023 account for 85% of young college graduate employment. The occupations with employment losses over the last three years were more likely to employ young noncollege workers than college graduates. It doesn’t appear that the occupations where young college graduates tend to work have been hit particularly hard in the last couple of years.

Figure B

While there has been job growth among occupations that tend to be filled by young college graduates, some worry about an increase in labor market underutilization, i.e., when workers with a college degree wind up working in jobs that typically don’t require one. Using O*NET data2, the New York Federal Reserve tracks this type of labor market underutilization. While the share of recent college graduates working at a job that doesn’t require a college degree has ticked up slightly over the last three years, it remains lower than it was for workers who graduated in the aftermath of the Great Recession. Even as late as 2017, young college graduates were working at these noncollege jobs at higher rates than they are today.

While college-educated workers are in more AI-exposed occupations, this does not appear to be driving labor market weakness

Much has been written in the last few years about AI exposure and its impact on the labor market. Using data from ADP, a large payroll processing company, Brynjolfsson, Chandar, and Chen find that entry-level workers in AI-exposed occupations—particularly AI uses that automate, not augment their work—have experienced an employment decline larger than that of older workers in the same occupations and all workers in less exposed occupations, explaining some of their stagnant overall employment growth. Atkinson and Yamco also find that declines in AI-exposed occupations are tied to lack of hiring rather than layoffs, hitting harder for young people attempting to enter the labor market. The second blog post in our series noted an across-the-board slowdown in hiring—which hurts the job prospects of all young workers, not only those in the industries most affected by AI.

On the other hand, researchers at the Yale Budget Lab argue that there has only been a slight increase in the shift in the occupation mix of employment, which would be evidence of AI automating jobs. They find that high AI-exposed occupations—determined by the top quintile of AI exposure—have yet to show declining employment, so no “dissimilarity” between young and older college graduates in terms of occupation mix has materialized. Raderman also finds that there isn’t strong evidence that AI is responsible for weaker labor market outcomes for recent college graduates, using evidence from Tillerman on college majors paired with change in unemployment.

Given the variation in assessments, we wanted to take a look at the data ourselves. Gimbel, Kendall, and Kulsakdinun have done an admirable job of summarizing the literature that attempts to classify AI exposure and propose a weighted aggregate measure of AI exposure.3 We employ this measure to investigate whether young college graduates may be more likely to be at risk in AI-exposed occupations than other workers.

In Figure C, we display the AI exposure of occupations weighted by the share of the entire workforce in each occupation. Moving from the left to the right on the figure increases AI intensity. For instance, professional and office & administrative support occupations are more AI exposed (to the right), while production, transportation, and service occupations are less AI exposed (to the left). Overall, the mean AI exposure score is 0.23.4

Figure C

In Figure D, we show the distribution of select demographic groups by occupation and AI exposure. As with earlier analysis, we compare young college graduates with all college graduates and young noncollege workers, in separate panels in the figure.

According to the aggregate measure, college graduates do have higher AI exposure in the labor market than the overall workforce. It is clear there is more mass in the direction of higher exposure (to the right) and their mean exposure is 1.07, higher than that of workers writ large. But the AI exposure of young college graduates isn’t any higher than that of college graduates in general. Mean AI exposure among young college graduates is 1.00.

Figure D

What is striking is that the AI exposure among young college graduates (1.00) is considerably higher than that of young noncollege workers (-0.61). If AI was driving labor market outcomes, we’d expect young college graduates to fare worse in today’s economy, e.g., see larger declines in employment or faster increases in unemployment. But, when we compare unemployment rates as we did in the first blog post of this series, both groups experienced similar increases in unemployment over the last two to three years. Trends in employment rates were also consistent across these groups.

Since the weakening labor market is hitting both young college and noncollege workers alike, it’s hard to argue that AI is uniquely causing job losses for new labor market entrants graduating from college now or in recent years. These findings are consistent with the literature, as there is currently no consensus about the effects of working in AI-exposed occupations on employment thus far.

1. Throughout this brief, we define young college graduates as people between the ages of 22 and 27 with only a four-year college degree. Unlike similar analyses of young workers, we do not exclude young college graduates who are currently enrolled in school, but the results here are robust either way. Unless otherwise noted, data for 2026 represent a 12-month average from April 2025 through March 2026 for the most up to date and reliable estimates, which removes seasonality and increases sample sizes.

2. O*NET or the Occupational Information Network provides the largest up-to-date database of information about workers sorted into detailed occupations. Information provided is about skills, abilities, education, training, and more.

3. We use an updated summary AI exposure PCA score (principal component analysis weighted standardized z-score) provided by the authors, May 13, 2026.

4. The PCA score scale is centered at 0, the unweighted mean across occupations.

Key takeaways

• The depressed overall hires rate is a key driver of new labor market weakness for young college graduates, as it makes it harder for them to break into the labor market. This is true across industries, not just in those that disproportionately employ young college graduates—suggesting the culprit is not a structural change in the economy like AI but a labor market in which employers are hiring less and workers are holding onto the jobs they have.
• The information sector—posited to be more AI-exposed—has experienced recent job losses but employs only 2.3% of young college graduates.
• High-tech industries, which employ about 1 in 10 college workers, expanded at a historically rapid pace in the early 2020s but have shown signs of softening over the last three years.

The early 2020s labor market for young college graduates was strong. But, as we showed in this series’ first blog post, the Class of 2026 is graduating college into a labor market that has notably weakened in the past two years. A growing share of young college graduates are looking for jobs, but their employment rates have not kept pace—meaning unemployment is rising faster for young graduates than for the overall workforce. While their outcomes remain better than those of their noncollege counterparts, the uptick in unemployment has been a rising concern.

In this blog post, we delve deeper into the industries where young college graduates are likely to work,1 examining whether it has been relatively more difficult to secure employment in these fields as the labor market has weakened. Our analysis first examines employment changes, then turns to labor market flows, including hires and separations rates. We also scour the data for signs of contraction in the tech sector that may disproportionately affect the prospects for young college graduates as they enter the labor market.

In the third blog post in the series, we will examine the occupations where young college graduates work with particular attention to occupations that may have grown or shrunk, as well as to those most exposed to AI.

Young college graduates work in industries with strong growth in this business cycle

Over half of young college graduates work in private education and health services, professional and business services, or public-sector jobs. Figure A displays the share of employment in each industrial sector or sector grouping for young college graduates ages 22 to 27, all college graduates, and young workers without a four-year college degree. Industries in the figure appear in order of the share of young college graduates they employ, from largest to smallest.

The types of jobs where young college graduates work look similar to those of college graduates generally. Young workers without a college degree (i.e., noncollege) are far more likely to work in trade, transportation, and utilities; mining, construction, and manufacturing; or leisure and hospitality. All groups of workers are least likely to work in the information sector, closely watched for signs of AI-induced displacement.

Figure A

Figure B shows the change in employment in each sector between 2019 and 2026 and between 2023 and 2026, arranged in the same order as Figure A for comparison. The two fastest growing sectors since the last business cycle peak occurred in the two largest sectors for young college graduates: private education and health services and professional and business services.

Since the rollout of ChatGPT, many have looked at industries and occupations likely to be exposed to AI to see whether this has led to weaker job growth. Among the most closely watched of these industries is the information sector, which has seen an 8.5% employment decline since 2023.2 While these losses are notable—and especially relevant to understanding AI’s fingerprints on the labor market for young college-educated workers—it cannot be overemphasized just how small this sector is in the overall economy. Less than 2% of overall employment is in the information sector, including only 2.3% of young college graduates. Further, the sector saw a rapid employment expansion between 2019 and 2023—the employment loss between 2019 and today is just 2.0%.

Figure B

High-tech sectors have relatively more college graduates but haven’t experienced large AI-induced employment losses

In recent years, the Census Bureau has created an experimental data series on high-tech industries to better understand business dynamics. These include both manufacturing and service sector components of high tech.3 We translate their NAICS classification for high-tech industries into Census Industry Classifications used by the Current Population Survey (CPS) to determine the likelihood of young college graduates working in these sectors.

In the 2026 economy, about 5.6% of workers were employed in what the Census considers high-tech industries. Just about 1 in 10 (9.9%) of the college-educated workforce works in the tech industry. Young college graduates are similarly represented: 10.3% work in tech.

Overall, the Current Establishment Survey tells us that tech industry employment tracked changes in overall private employment in the prior business cycle (between 2007 and 2019) but expanded sharply in the early 2020s and has softened a bit in the last three years. Since 2023, the tech sector has fallen by 0.7%. While overall employment using CPS does show modest growth, neither shows large swings that suggest a large impact for young college graduates.4

Weak hires may be the biggest culprit to labor market weakness for young college graduates

The Job Openings and Labor Turnover Survey (JOLTS) can shed light on the question of whether entry-level workers—with or without college degrees—are facing a harder labor market to break into. While JOLTS doesn’t include demographic characteristics, it presents jobs openings as well as rates of hiring, layoffs, quits, and other separations. Today’s economy has substantially less churn than during the recovery from the pandemic, when millions of workers reentered the labor market after mass layoffs—many quit soon after as they searched for, and generally found, better opportunities.

Figure C shows the hires and separation rates. The lighter colors represent the monthly seasonally adjusted data for each series while the darker colors represent a 12-month moving average that provides a better overall picture of recent trends, smoothing out some data volatility. Over the last five years, the hires rate has steadily fallen and now sits at levels last seen in 2013 and 2014, when the labor market was still struggling to recover from the Great Recession.

The total separations rate includes quits, layoffs and discharges, and other separations.5 As with the hires rate, the separations rate has been declining over the last few years and now sits about where it was in 2014. Much of this is due to reductions in quits. Quits are higher when workers feel confident that they will find better job opportunities. Right now, workers are sitting tight, more so than any point in the past 10 years. Taken together, there is simply less churn in the labor market. But reduced churn is not inherently bad. If the frantic labor market of the early 2020s led to many workers and employers finding satisfactory matches, it could make sense that the following years would see less churn than normal. But for young workers looking to enter the job market, a reduction in hiring can make it harder to find a job.

Figure C

Table 1 breaks down the change in the hires and separations rate over the last three years, again using 12-month moving averages to smooth some volatility in the data. The industries are listed in order of the share of young college graduates they employ, similarly to Figure A. Overall, the hires rate fell 0.8 percentage points and the separations rate fell 0.6 percentage points between 2023 and 2026. The industries where young college graduates are more likely to work saw smaller reductions in both hires and quits than the overall. Industries where young workers without a college degree are more often found—over a quarter are in trade, transportation, and utilities—saw greater losses. Finally, leisure and hospitality, where young noncollege are more than twice as likely to work as young college graduates, saw the largest declines in hiring.

It does not appear that the industries where young college graduates tend to work are experiencing more weakness than other industries. Job gains are just as strong, if not stronger, and hiring hasn’t fallen as far in other industries. In short, there does not seem to be any profound structural change in the economy affecting the industry composition of employment—AI or anything else—that would easily explain the softening of the labor market for young college graduates in recent years. What it does appear to be is a harder labor market for young workers to break into when employers are less likely to hire and workers are more likely to sit tight in the job they have.

1. Throughout this blog post, we define young college graduates as people between the ages of 22 and 27 with only a four-year college degree. Unlike similar analyses of young workers, we do not exclude young college graduates that are currently enrolled in school, but the results here are robust either way. Unless otherwise noted, data for 2026 represent a 12-month average from April 2025 through March 2026 for the most up to date and reliable estimates, which removes seasonality and increases sample sizes.

2. ChatGPT was first introduced in November 2022 but took several months for more widespread usage.

3. High tech industries: Computer and Peripheral Equipment Manufacturing, Communications Equipment Manufacturing, Semiconductor and Other Electronic Component Manufacturing, Navigational, Measuring, Electromedical, and Control Instruments Manufacturing, Aerospace Product and Parts Manufacturing, Software Publishers, Data Processing, Hosting, and Related Services, Other Information Services, Architectural, Engineering, and Related Services, Computer Systems Design and Related Services, and Scientific Research and Development Services.

4.  It’s not unusual for the CPS and CES to display small differences in employment levels or trends considering nontrivial differences in their methodologies. 

5. Other separations include separations due to retirement, death, disability, and transfers to other locations of the same firm.

Since the second Trump administration swept into office in January 2025, it has undertaken a range of damaging and destabilizing actions that have weakened the economy, undermined workers, hurt businesses and consumers, and threatened core elements of our democracy. While Trump has targeted numerous Democratic-led states and cities, the Washington, D.C., region has faced acute and prolonged harms since day one. From the first set of executive actions signed on Inauguration Day, the Trump administration has attacked people and businesses in the capital region repeatedly and intensely. These initial actions announced the president’s dubious claims of authority to fire large segments of the federal workforce, eliminate long-standing federal agencies and programs, and begin a campaign of illegal and inhumane mass deportations.  

The Trump administration’s damaging actions have been enabled and abetted by Republican members of Congress. Their passage of H.R. 1, the bill that the White House has referred to as the “One Big Beautiful Bill Act” (OBBBA), amplifies the administration’s mass deportation agenda and shreds critical health care and food supports for lower-income families to finance tax cuts for the wealthy. This funding bill will only cause more pain in the years ahead for Washington, D.C.-area households and throughout the country.

Congress also passed a federal spending bill that constrained the District of Columbia’s ability to spend its own tax revenue (Koma 2025) and a resolution that may force the district to adopt local tax code changes that match the OBBBA, whether the city wants to or not—changes that will jeopardize hundreds of millions of dollars for city programs (D.C. Fiscal Policy Institute 2026).

In this report, we assess the early indicators of the damage of Trump’s actions and their effects on the Washington, D.C., regional economy, with particular attention to effects on workers and the labor market. We focus on this region due to its prominence as an early target of the Trump administration, in part due to its large federal workforce. Additionally, the district’s unique status as a non-state means that its leaders have far less legal authority to resist Trump’s interference than other target areas do.

Throughout this report, unless otherwise indicated, the data describe economic conditions for the Washington, D.C., metropolitan statistical area (MSA), which includes the District of Columbia, four nearby counties in Maryland, six cities and 11 counties in northern Virginia, and one county in West Virginia. We also refer to this region as the DMV (Washington, D.C.; Maryland; and Virginia). While we do not yet have the requisite data to fully and precisely document all the effects of the administration’s actions, we can see clear signals that the regional economy is already struggling, with more severe impacts likely to register in the data soon.

We then explore some of the factors that make other regions particularly vulnerable to significant economic harm from the Trump administration’s agenda. These include counties with large concentrations of federal workers, areas where federal transfer income (such as Medicaid and Social Security) makes up a significant portion of the region’s economic base, and places with significant immigrant populations. Though Trump has largely targeted prominent, Democratic-led areas, many of the regions most susceptible to the harmful economic consequences of the administration’s actions are rural counties, frequently represented in Congress by Republicans.

Trump’s actions in Washington, D.C., have led to reduced employment and rising unemployment

The clearest sign of the harm that the Trump administration’s actions have done to the Washington, D.C., regional economy is the substantial drop in the region’s employment rate. Based on EPI analysis of Current Population Survey data from the Bureau of Labor Statistics, from December 2024 to December 2025, the share of the regional working-age population with a job fell by 3.2 percentage points.1 As shown in Table 1, this compares with a decline of just 0.4 percentage points for the country over the same period. Among prime-age workers (those ages 25–54), the share employed in the DMV fell by 2.7 percentage points, compared with a decline of just 0.1 percentage points for the country overall.

This dramatic drop in regional employment is a direct result of the Trump administration’s relentless attacks on federal government workers, cuts to federal programs and agencies, and their cascading effects on connected regional industries. Prior to Trump’s taking office, federal employees made up 11.2% of the metro area’s total workforce (BLS-CES-SAE 2025).2 Between the end of 2024 and 2025, federal employment in the DMV region fell by more than 53,800 jobs (-14.2%) (BLS-CES-SAE 2026).3 These losses reverberated through the regional economy as affected households pulled back on spending, and many may have even opted to move, as data show the DMV region had the largest increase in home sale listings of any major metro last year (Brookings Institution 2026).

These significant cuts to federal employment, though highly damaging on their own, are only the first layer of the administration’s harm on the regional labor market. The DMV has a non-federal workforce of over three million people (BLS-CES-SAE 2026), many of whom work at firms that consult with, contract with, are funded by, or are otherwise connected to the government.4 The Trump administration has terminated thousands of grants to scientific research institutions (Kozlov, Tollefson, and Garisto 2026) and frozen or delayed funding for tens of thousands of nonprofit organizations, causing those targeted to limit operations or lay off staff (Tomasko et al. 2025). These cuts have also shrunk the funding pool for nonprofit groups, causing budget challenges even for those not previously receiving federal funding, as they must compete with groups previously funded through federal programs that are now scrambling to fill gaps with private support (Barrett 2025). The administration has also moved to cancel contracts with any company that maintains a commitment to DEI standards (Singh 2026). Although these cuts affect organizations everywhere, the DMV is disproportionately vulnerable to the economic harms of attacks on this sector as it has one of the highest concentrations of nonprofits in the country (Friesenhahn 2025). This is evident in the region’s slight dip (-0.3%) in private-sector employment from December 2024 to December 2025, a change from the consistent, albeit slowing, growth that had marked the years following the COVID-19 pandemic. At the national level, private-sector employment experienced slow but still positive change (0.5%) over the same period (BLS-CES-SAE 2026).5

The widespread impact of the administration’s actions can be seen in the breadth of employment declines across racial, ethnic, gender, and age groups in the region. As shown in Table 1, the employment rate fell by at least 2 percentage points for every demographic category of workers in the DMV. Notably, young workers under age 25 (-4.3 percentage points), workers age 55 and older (-3.3 percentage points), men (-3.5 percentage points), and Black workers (-5.9 percentage points) all experienced drops in their employment rates larger than the regional average. For older workers, the above-average decline likely reflects, at least in part, the firings and retirements of many federal employees, including many who had been near retirement age and opted into the so-called “Fork in the Road” deferred resignation program. For young workers, the administration’s funding and programmatic cuts directly reduced many traditional Beltway early-career opportunities (internships, fellowships), while weakness in the broader regional economy simultaneously forced area employers to pull back on entry-level positions.

Still, not all groups have been equally affected by Trump’s actions. As Table 1 shows, Black workers in the DMV region have suffered the largest employment declines, with the share employed falling by 5.9 percentage points in 2025. This is nearly triple the employment drop experienced by white workers (2.0 percentage points) in the region and, notably, more than seven times the employment drop of Black workers throughout the country overall (0.8 percentage points). Again, this is a direct consequence of the administration’s attacks on the federal workforce. Black workers have long tended to make up a larger share of the public sector than they do in the private sector—both in the DMV and across the country. This is because the public sector has historically been a pathway to the middle class for workers of color who face labor market discrimination in the private sector (Maye and Marvin 2025).

Trump’s massive cuts to federal employment have also rapidly undone what had been considerable progress in shrinking the regional Black-white employment gap. Figure A shows the employment rate of DMV workers, overall and by race/ethnicity, since the end of 2018. The rapid drop in the Black employment rate since the start of President Trump’s second term is striking, bringing the regional Black employment rate back down to its pandemic-era low. It is also notable that before that drop began, Black workers in the region were employed at essentially the same rate as their white counterparts—the only time in the last two decades when that occurred. These losses in employment will exacerbate existing racial and gender inequity across wages, poverty, and unemployment (Markoff and Zielinski 2026; Zielinski 2025; Busette and Elizondo 2022).

Recent increases in the DMV’s overall unemployment rate underscore the damage Trump is doing to the region. The non-seasonally adjusted unemployment rate jumped more than a full percentage point, from 3.1% in January 2025 to 4.4% in January 2026—more than four times the increase in the national figure. (Importantly, this increase understates the weakening of the area labor market, as the BLS estimates the DMV labor force shrank by 3% over the same period—meaning that many workers who would have been counted as unemployed simply left the area labor force.) For comparison, the national non-seasonally adjusted unemployment rate increased by less than half a percentage point, moving from 4.4% in January 2025 to 4.7% in January 2026 (BLS-LAUS 2026).

These numbers do not capture the full extent of the economic downturn in the DMV area, nor can they give us precise insight into where the pain has been most acutely felt. The administration’s violent deportation agenda, for example, will lead to a drop in immigrant and U.S.-born Hispanic workers’ employment, but resulting changes in Hispanic employment rates may be muted by the corresponding shrinking of the overall Hispanic population (Zipperer 2025). In other words, while the overall Hispanic population in the U.S. may fall dramatically in coming years, the ratio of remaining employed workers to remaining total population may stay somewhat consistent. This will mask the true scale of the economic and social harm being done to immigrant communities in the DMV and across the country.

It is also difficult to fully quantify how the deployment and continued presence of National Guard troops, violent immigration actions, and other authoritarian, fear-inducing tactics have impacted D.C.-area businesses, workers, and families, particularly in neighborhoods with predominately Black and Latino populations. Early data show regional declines in tourism, consumer spending, and foot traffic; harder to capture are the emotional and long-term economic consequences (Montgomery 2025; Hadden Loh and Haskins 2025; Sachs and Cocco 2025). Other recent analyses estimate similar economic harms in cities where targeted federal immigration enforcement actions have been aggressively deployed (Rosenthal and Sojourner 2026). A full accounting of the Trump administration’s harms on the Washington, D.C., region will take years to document.

Other localities should brace for similar consequences

Some of the Trump administration’s actions and their acute consequences are unique to the DMV, a function of the region’s high concentration of federal employees and government contractors, as well as the District of Columbia’s lack of statehood and full constitutional rights. However, the anti-government attacks the administration has unleashed on DMV-area households, workers, and businesses will have cascading consequences for communities throughout the country. The effects of the administration’s authoritarian attacks on the civil service, democratic institutions, and immigrants (Human Rights Watch 2026) that first registered across the DMV should be viewed as a preview of the consequences that will be felt in other regions. While no locality will be spared, regions particularly at risk include those with large shares of government workers (especially federal workers, but state and local government workers too), localities in which federal funding and social safety net programs make up a large portion of total area income, and those with large immigrant populations.

Trump’s attacks on the federal workforce will harm communities that rely on their employment

The day Trump returned to power in January 2025, he began attacking the federal workforce, first by moving to reclassify tens of thousands of federal employees to make it easier to fire and replace them with political loyalists (EPI 2026c), and then by stripping more than one million federal workers of their collective bargaining rights (EPI 2025a). The Trump White House subsequently worked feverishly to slash federal employment, attempting large and chaotic reductions in force, shuttering entire agencies, and coercing tens of thousands of staff to resign, among many other attacks (Poydock 2025). As of March 2026, the administration’s actions have reduced nationwide federal government employment by over 350,000 (11.7%) since January 2025 (Gould 2026).

Though federal workers make up a sizeable share of the DMV’s workforce, over 80% of federal workers live outside the region (Partnership for Public Service 2024). For instance, in Alaska, Hawaii, and New Mexico—states that are home to large swaths of federal and Native land, military bases, and federal research institutions—federal workers make up at least 4.5% of total employment (EPI 2025c). Within states, federal workers tend to be concentrated in specific localities. For instance, in Apache County, Arizona, which is largely made up of the Navajo Nation and the White Mountain Apache Reservations, lands that extend beyond county lines, the federal government employs 12% of the county’s workers, more than double the next most significant county for federal worker employment in the state (EPI 2025c). There are 22 U.S. counties, spread across the South, Midwest, and West Census regions, where federal workers comprise at least 10% of the county’s workforce (see Table 2).

In these counties and elsewhere, federal workers are the backbone of the regional economy, both through the essential services they provide and through their contributions to the local economy. Trump’s attacks simultaneously threaten federal workers’ livelihoods and the economic health of communities in which these workers’ spending on goods and services makes up a large share of economic activity in the region. In Apache County, Arizona, civilian government workers’ earnings comprise 11.7% of total economic activity in the county (see Table 3)—roughly the same as their share of overall county employment. However, in some counties, federal employees’ earnings are a disproportionate share of the regional economic base. For instance, in Leavenworth County, Kansas, where federal employees make up 10.0% of employment (Leavenworth has a large federal prison), federal civilian earnings comprise 22.1% of total income in the county.

The effects from lost federal jobs and income in these regions could be devastating. Some of these communities are places that have already faced historic disinvestment and in which there are few local employment opportunities that can match the quality of federal government jobs. These jobs are historically stable, good quality, union jobs that offer a pathway to the middle class, particularly for workers without a college education.6

Regions highly dependent on federal revenue will also suffer from a reduction in services and a loss of income

Beyond the harm to localities from reductions in the federal workforce, localities that are particularly reliant on federal government revenue and services will bear the consequences of Trump’s actions most acutely, though no locality will be spared from harm. For example, the Trump administration has announced or considered $23 billion in cuts to federal clean energy projects in nearly every state (CATF 2025) and $8 billion in cuts to colleges and universities that will impact every state’s economy (Bedekovics and Ragland 2025). Trump’s 2025 budget bill also made massive cuts to federal safety net programs that millions of low-income households rely on in order to finance tax cuts for the wealthiest households and corporations.

Funds from federal programs such as SNAP, Medicaid, and other social programs not only help struggling families make ends meet, they also comprise a significant share of a locality’s “economic base,” the amount of money circulating in that region, as shown by sociologist Robert Manduca in a recent working paper (2025). Indeed, an often-overlooked benefit of Medicaid coverage is its role as a source of income for low-income households (money they would have had to spend on medical care in the absence of Medicaid). For the bottom 20% of households in the U.S., Medicaid comprised 70% of their total money income, based on recent data from the Congressional Budget Office (Bivens, Wething, and Morrissey 2025). In fact, government transfers such as Social Security, Medicare, and Medicaid collectively made up 40% of the economic base of U.S. regions in 2022 (Manduca 2025). Substantial cuts to government social programs that support low-income households could reduce the economic base of these localities, at a scale equivalent, in many cases, to the loss of entire private industries in those areas.

Without deliberate intervention by state lawmakers to offset lost federal revenues, localities in every state face dire economic losses, but states particularly reliant on government transfers will suffer most. For instance, take Clay County, West Virginia, which is represented in Congress by Rep. Carol Miller (R-WV01), who voted in support of Trump’s budget bill (Miller 2025). Clay County’s poverty rate is more than double the national rate, and its per capita income is half the national amount (U.S. Census 2024a). Of the 10 U.S. counties that rely most on each of the largest federal social insurance programs (Medicare, Medicaid, SNAP, and Social Security) as a share of their economic base, Clay is the only county in the country to show up three times (see Table 4). Federal government transfers in the form of Medicare, SNAP, and Social Security payments comprise 57% of Clay County’s economic base, 20 times the share comprised by the earnings of every private industry in the county combined. Alaska, Arizona, Florida, Georgia, Kentucky, Tennessee, and West Virginia all have at least three counties that are ranked in the top 10 in the country for their reliance on a given social safety net program as a share of the county’s economic base (see Table 4).

Localities that have significant shares of federal workers and rely heavily on federal government transfers may face particularly significant consequences as a result of Trump’s attacks on the federal workforce and the Republican budget bill’s cuts to essential social safety net programs. For example, in Rio Arriba County, New Mexico, and Apache County, Arizona, federal government workers make up 16.1% and 12.0% of all workers in the county, respectively (EPI 2025b). At the same time, both counties are ranked in the top-10 counties most reliant on federal government transfers—Apache is #2 for Medicaid, and Rio Arriba is #10 for SNAP. In Apache County, federal government transfers account for three-quarters (76.9%) of the county’s economic base, and the earnings of federal government civilian workers account for 11.7%—the Navajo Nation Tribal Government is the county’s largest employer (NACOG 2023). Meanwhile, private earnings account for a mere 2.8% of the county’s economy. In Apache, Trump’s cuts to both the federal workforce and federal government programs mean that the federal government may be unable to fulfill its legal obligations to tribal communities (Brown 2025) that have faced decades of disinvestment and depressed economic outcomes resulting from historic land theft and forced assimilation. Apache County’s poverty rate of 31.2% (AZ Economics 2026) is nearly triple the national rate of 11.1% in 2023 (Shrider 2024).

Trump’s anti-immigrant crackdown and deportation agenda hurt localities with large immigrant populations

Trump has launched a campaign of terror against immigrant communities, communities of color, and those who stand with them. Last summer, Trump federalized local police and deployed thousands of federal troops to diverse cities with large immigrant populations (Kim 2025). Though Washington, D.C., may have experienced the most visible federal troop presence, a function of the district’s lack of statehood and the president’s unchecked authority to mobilize the National Guard there (Dallas 2025), Los Angeles was the first city Trump targeted after public opposition to aggressive immigration raids (Kim 2025). It was soon followed by Washington, D.C.; Memphis, Tennessee; Portland, Oregon; New Orleans, Louisiana; Minneapolis, Minnesota; and Portland, Maine.

These attacks are characteristic of an authoritarian playbook that includes forcing the leaders of diverse, opposition-led communities to bend to the strongman government’s will (McManus, Benson, and Herman 2024). Minneapolis, home to a large immigrant population, was subjected to an unprecedented immigration crackdown that drew widespread protests (Boone 2026). During “Operation Metro Surge,” as it was called, federal immigration enforcement officials made 4,000 arrests and killed two U.S. citizens. Though the true toll of this violent operation may never be fully quantified, initial economic data show clear cause for concern. A recent analysis estimated that Trump’s immigration crackdown has led to a 2.9% decline in consumer spending in Minnesota over a single month—the equivalent of the state’s economy losing $626 million (Rosenthal and Sojourner 2026). Relative to overall consumer spending, the food and accommodation sector (which employs a large share of immigrant workers) saw the most significant decline in January 2026—3.8% or a $46 million reduction in economic activity. Researchers also estimated that nearly 3% of workers in the Minneapolis-Saint Paul region were unable to work during the occupation, resulting in a loss of over $100 million in wages (Sojourner and Rosenthal 2026).

Trump’s deportation agenda will continue to destabilize local communities and result in job losses for immigrant and U.S.-born residents alike (Zipperer 2025). Though immigrants live in counties across the U.S., coastal urban areas tend to have the largest shares of foreign-born residents. Counties with the largest foreign-born populations include Miami-Dade, Florida; Queens, New York; Aleutians, Alaska; and Hudson, New Jersey (see Table 5). Counties with relatively large shares of immigrants may see particularly acute harms from aggressive immigration enforcement.

Communities face overlapping economic threats from attacks on federal workers, the social safety net, and immigrants, but state and local lawmakers can resist them.

The Trump administration’s attacks on the federal workforce, the social safety net, and immigrant communities are designed to exacerbate economic precarity in many communities that are already struggling (Bivens 2026). The implementation of Trump’s authoritarian agenda in the DMV region may be the first, clearest, and in some cases most direct manifestation of its harms, but other localities across the country—particularly those with large federal workforces, those that are heavily dependent on federal revenue and those with sizeable immigrant populations—are far from immune, and many will suffer as much, if not more, from this agenda.

While state and local leaders cannot stop federal attacks, they do have the power to resist Trump’s agenda by improving state labor standards (EPI 2026b), advancing protections for immigrant workers (Díaz and Whitaker 2026), investing in the public-sector workforce (Bivens and Shierholz 2026), and using progressive tax policies (Austin and Davis 2025) to stabilize funding for critical social programs and other investments that workers, families, and communities need.

Notes

1. Throughout this report, unless explicitly noted, the source for all employment rate data is the authors’ analysis of Current Population Survey data (EPI 2026a). We compare an average of calendar year 2025 with calendar year 2024 in order to have adequate sample sizes for the noted demographic groups.

2. Employment level by industry and sector data come from the authors’ analysis of the Bureau of Labor Statistics’ Current Employment Statistics (CES) State and Metro Area (SAE) data.

3. These numbers are calculated using monthly totals rather than annual averages. A quarterly comparison of 2025Q4 to 2024Q4 finds roughly the same results—employment fell by 52,600 jobs (13.9%). The quarterly analysis omits October in both years to maintain an apples-to-apples comparison, accounting for missing data due to the government shutdown that began in October 2025 and the subsequent lapse in Bureau of Labor Statistics funding.

4. The non-federal workforce includes private sector workers as well as state and local government employees.

5. These numbers are calculated using monthly totals rather than annual averages. Quarterly comparisons of 2025 Q4 to 2024 Q4 produce similar results—private sector employment fell by 0.1% in the DMV and grew by 0.7% nationally. The quarterly analysis follows the methodology outlined in note 2.

6. On average, federal workers with advanced degrees typically earn less in wages and total compensation than their private-sector counterparts. Federal workers without an advanced degree typically earn more than their private-sector counterparts and have access to retirement benefits that have become less common in the private sector (CBO 2024).

References

Austin, Sarah, and Carl Davis. 2025. The Wealth Proceeds Tax: A Simple Way for States to Tax the Wealthy. Institute on Taxation and Economic Policy, October 2025.

AZ Economics. 2026 “Apache County, Arizona.” U.S. Economic Research. Accessed April 2026.

Barrett, William P. 2025. “America’s Top 100 Charities: A Year of Pain After Trump Cuts.” Forbes, December 12, 2025.

Bedekovics, Gréta, and Will Ragland. 2025. Mapping Federal Funding Cuts to U.S. Colleges and Universities. Center for American Progress, July 2025.

Bivens, Josh. 2026. The Trump Administration’s Macroeconomic Agenda Harms Affordability and Raises Inequality. Economic Policy Institute, February 2026.

Bivens, Josh, and Heidi Shierholz. 2026. “You Can’t Starve the Public Sector to Excellence.” Working Economics Blog (Economic Policy Institute), February 27, 2026.

Bivens, Josh, Hilary Wething, and Monique Morrissey. 2025. Cutting Medicaid to Pay for Low Taxes on the Rich Is a Terrible Trade for American Families. Economic Policy Institute, February 2025.

Boone, Rebecca. 2026. “A Timeline of Trump’s Immigration Crackdown in Minnesota.” PBS Newshour, February 13, 2026.

Brookings Institution. 2026. “Active Residential For-Sale Listings,” DMV Monitor. Last updated February 18, 2026.

Brown, Alex. 2025. “For Indian Country, Federal Cuts Decimate Core Tribal Programs.” Stateline, March 4, 2025.

Bureau of Labor Statistics, Current Employment Statistics State and Metro Area (BLS-CES-SAE). Various years. Public data series accessed through the CES State and Metro Area Databases and through series reports. Accessed April 2026.

Bureau of Labor Statistics, Local Area Unemployment Statistics (BLS-LAUS). Various years. Data from the LAUS are available through the LAUS database and through series reports. Accessed April 2026.

Busette, Camille, and Samantha Elizondo. 2022. “Economic Disparities in the Washington, D.C. Metro Region Provide Opportunities for Policy Action.” Commentary, Brookings Institution, April 27, 2022.

Clean Air Task Force (CATF). 2025. “The High Cost of Retreat: Impacts of Department of Energy Project Cuts.” Clean Air Task Force, November 21, 2025.

Congressional Budget Office (CBO). 2024. Comparing the Compensation of Federal and Private-Sector Employees in 2022. Congressional Budget Office, April 2024.

Dallas, Kelsey. 2025. “The President’s Power to Deploy Troops Domestically: An Explainer.” SCOTUSblog, October 28, 2025.

D.C. Fiscal Policy Institute. 2026. “Congressional Interference Will Cost D.C. Nearly $700 Million in Local Revenue and Jeopardize Efforts to Reduce Child Poverty.” D.C. Fiscal Policy Institute, February 4, 2026.

Díaz, Marisa, and Mimi Whitaker. 2026. How States and Localities Can Strengthen Workplace Protections for Immigrant Workers. National Employment Law Project, January 2026.

Economic Policy Institute (EPI). 2025a. “Executive Order on ‘Exclusions from Federal Labor-Management Relations Programs.’” Federal Policy Watch (Economic Policy Institute), December 17, 2025.

Economic Policy Institute (EPI). 2025b. How Many Federal Employees Live in Your State? Economic Policy Institute.

Economic Policy Institute (EPI). 2025c. “New Resource Calculates How Many Federal Workers Live in Every State, County, and Congressional District” Economic Policy Institute (press release). March 3, 2025.

Economic Policy Institute (EPI). 2026a. Current Population Survey Extracts, Version 2026.3.11, https://microdata.epi.org.

Economic Policy Institute (EPI). 2026b. Holding the Line: State Solutions to the U.S. Worker Rights Crisis. Economic Policy Institute.

Economic Policy Institute (EPI). 2026c. “OPM Finalizes Regulation Enabling Firing Federal Employees for Political Reasons.” Federal Policy Watch (Economic Policy Institute), March 4, 2026.

Friesenhahn, Erik. 2025. “Nonprofit Organizations: State and Regional Employment Trends.” Monthly Labor Review (U.S. Bureau of Labor Statistics), March 2025. https://doi.org/10.21916/mlr.2025.6.

Gould, Elise. 2026. “Attacks on the federal workforce continue (down 18k jobs in March).” Bluesky, @elisegould.bluesky.social, April 3, 2026, 9:01 a.m.

Hadden Loh, Tracy, and Glencora Haskins. 2025. Consumer Spending and Visitor Demand in the Washington, D.C. Region Are Dropping. Brookings Institution, December 2025.

Human Rights Watch. 2026. “Sliding Towards Authoritarianism?” January 2026.

Kim, Juliana. 2025. “Trump Says National Guard Will Soon Go to New Orleans. Here’s the Latest.” NPR, December 3, 2025.

Koma, Alex. 2025. “Here’s How D.C. Solved the Billion-Dollar Budget Problem Congress Created.” WAMU, October 22, 2025.

Kozlov, Max, Jeff Tollefson, and Dan Garisto. 2026. “U.S. Science After a Year of Trump.” Nature 649 (January): 812–815.

Lynch, Teresa M., and Robert Manduca. 2024. “Beyond Local and Traded: Evidence for a Third Industry Market Area Type and Implications for Regional Economic Development.” Economic Development Quarterly 38, no. 3: 183–194, July 2024. 

Manduca, Robert. 2025. Financial and Transfer Income as Components of the Regional Economic Base. Washington Center for Equitable Growth, June 2025.

Markoff, Shira, and Connor Zielinski. 2026. Chronic Racial Inequality Holds Back Workers and Equitable Economic Growth. D.C. Fiscal Policy Institute, March 2026.

Maye, Adewale A., and Stevie Marvin. 2025. “Trump Attacks on Federal Agencies Have Steep Implications for Black Workers.” Working Economics Blog (Economic Policy Institute), April 10, 2025.

McManus, Allison, Robert Benson, and Dan Herman. 2024 “The Dangers of Project 2025: Global Lessons in Authoritarianism.” Center for American Progress, October 2024.

Miller, Carol. 2025. “Miller Votes to Send the One, Big, Beautiful Bill to President Trump’s Desk” (press release). Office of Congresswoman Carol Miller, West Virginia’s First District, July 3, 2025.

Montgomery, Mimi. 2025. “Trump Crackdown Is Affecting D.C.’s Image and Tourism Numbers.” Axios, August 29, 2025.

Northern Arizona Council of Governments (NACOG). 2023. “Business, Jobs, and Industry Highlights for Apache County.” Northern Arizona Council of Governments, November 20, 2023.

Partnership for Public Service. 2024. Beyond the Capital: The Federal Workforce Outside the D.C. Area. March 2024.

Poydock, Margaret. 2025. “How Trump Has Dismantled the Federal Workforce in His First 100 Days.” Working Economics Blog (Economic Policy Institute), May 23, 2025.

Rosenthal, Aaron, and Aaron Sojourner. 2026. The Economic Impact of Operation Metro Surge in January 2026: A Synthetic Difference-in-Differences Analysis. North Star Policy Action, February 2026.

Sachs, Andrea, and Federica Cocco. 2025. “D.C. Tourism Was Already Struggling. Then the National Guard Arrived.” Washington Post, August 29, 2025.

Shrider, Emily A. 2024. Poverty in the United States: 2023. United States Census Bureau, Report Number P60-283, September 2024.

Singh, Kanishka. 2026. “Trump Signs Executive Order Asking Federal Contractors to Eliminate DEI.” Reuters, March 26, 2026.

Sojourner, Aaron, and Aaron Rosenthal. 2026. Impact of DHS Agent Surge on Minneapolis-Saint Paul Metro Area Labor Outcomes. North Star Policy Action, February 2026.

Tomasko, Laura, Hannah Martin, Katie Fallon, Mirae Kim, Lewis Faulk, and Elizabeth T. Boris. 2025. How Government Funding Disruptions Affected Nonprofits in Early 2025: Nationally Representative Findings from the Nonprofit Trends and Impacts Study. Urban Institute, October 2025.

U.S. Census Bureau. 2024a. “American Community Survey 5-Year Estimates: Retrieved from Census Reporter Profile Page for Clay County, WV.” Accessed April 14, 2026.

U.S. Census Bureau. 2024b. “U.S. Foreign-Born Population: 2018–2022 American Community Survey, 5 Year-Estimates (Table B05006).” Accessed April 14, 2026.

Zielinski, Connor. 2025. “Inequality Remained Extreme in 2024 as D.C. Backslid on Poverty.” DCFPI Blog (D.C. Fiscal Policy Institute), September 15, 2025.

Zipperer, Ben. 2025. Trump’s Deportation Agenda Will Destroy Millions of Jobs: Both Immigrants and U.S.-Born Workers Would Suffer Lob losses, Particularly in Construction and Child Care. Economic Policy Institute, July 2025.

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

Comments on RIN 1235-AA46: Employee or Independent Contractor Status under the Fair Labor Standards Act, Family and Medical Leave Act, and Migrant and Seasonal Agricultural Worker Protection Act

Dear Director Navarrete:

We submit these comments from the Economic Policy Institute (EPI) on the Department of Labor’s (“Department” or “DOL”) Notice of Proposed Rulemaking (“NPRM”) regarding the standard for determining who is an employee and who is an independent contractor under the Fair Labor Standards Act (“FLSA”), the Family and Medical Leave Act (“FMLA”) and the Migrant and Seasonal Agricultural Worker Protection Act (“MSPA”).

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.

We strongly oppose the Department’s rule as proposed. We urge the Department to withdraw this rule and instead allow the long-standing test for determining employee status under the FLSA to stand.

EPI has conducted extensive research and analysis over the years on the harms of worker misclassification. As we have outlined, workers classified as independent contractors have no right to earn the federal minimum wage, or to earn overtime pay. They lose eligibility for unemployment insurance if they lose their work, and to workers’ compensation if they are injured on the job. They are less likely to receive employer-provided job benefits, such as health insurance and retirement benefits. They lose the right to paid sick or family leave in states and localities that extend those rights, and they would lose the right to even unpaid, but job-protected, family and medical leave under FMLA. Workers classified as independent contractors also must assume the full financial cost of Social Security and Medicare contributions, rather than split it evenly with their employer.

We attach here an April 2026 EPI report1 estimating the concrete economic costs of misclassification for 11 commonly misclassified types of jobs, among those most likely to be negatively affected by this rule. These include lower-wage, labor intensive jobs such as call center workers, landscaping workers, janitors and cleaners, home health aides, truck drivers, delivery workers, manicurists, housekeeping cleaners, retail sales workers, security guards, and construction workers. Workers in these and other occupations stand to lose wages, benefits, and the basic labor protections they should be owed under the FLSA.

The FLSA has a plain-language definition of “employ,” which “includes to suffer or permit to work.” This is a deliberately broad definition that was intended to provide the FLSA’s protections to most workers. The NPRM also seeks to once again upend the clear, long-standing “economic reality” test, which examines multiple factors to get to the central issue of worker classification: is the worker truly in business for themselves, or do they depend economically on finding work in the business of others, under the control and terms of the employer?

Instead of examining all of the relevant factors in a worker’s situation, the NPRM proposes elevating the factors of the level of control the employer exerts, and the worker’s opportunity for profit or loss, above all others in making a determination about whether someone is truly in business for themselves.

This would fail to account for the economic realities of many working relationships: for instance, would the primary work of the employer be able to get done without the worker? How permanent or exclusive is the work being performed—is there a fixed ending date? Does the worker invest in their own tools and equipment, marketing, or business plan, or is it the employer making those investments? Does the worker rely on the employer for training on how to get the job done? All of these questions fall under the factors that the NPRM would deprioritize—even though they provide important information about whether or not someone is truly in business for themselves, and thus that the employer doesn’t have an obligation to them under the FLSA.

This would narrow the definition of who is a covered employee under these three statutes. DOL’s NPRM will encourage misclassification schemes and a race to the bottom, where employers will be able to reclassify their employees as independent contractors and evade their obligations under these laws. Further, because of occupational segregation and other labor market disparities, people of color, women, and immigrants—and people at the intersections of these categories—are more likely to be in occupations where misclassification is common.

An analysis of the proposed rule’s potential costs to workers

In the proposed rule, the Department egregiously fails to estimate the transfers between employers, workers, and the social insurance system that would occur if this proposal were finalized. The requirements that agencies must follow as a part of the rulemaking process are very clear, and among them is the requirement that agencies must assess all quantifiable costs and benefits “to the fullest extent that these can be usefully estimated.”2 There is no question that DOL could have produced estimates; in what follows, we show that it is straightforward to produce estimates using data researchers routinely use and taking a methodological approach that is in the spirit of estimates the Department of Labor undertakes on a regular basis. One plausible explanation for why DOL left out the required estimate is that any good-faith estimate would have shown this rule will result in a substantial transfer from workers and the social insurance system to employers.

The Department only briefly touches on potential benefits to workers from their proposal. DOL estimates a 1-3% increase in the total number of independent contractors as a result of their proposed rule.3 However, DOL appears to assume that this increase will come entirely from people who were otherwise not engaging in paid work entering the workforce anew as independent contractors. This means the Department also assumes that there will not be significant reclassification of workers who are currently employees to independent contractor status. Given what we know about the scale of misclassification already occurring under current law, this seems to be, at best, a woefully naive understanding of what employers might do when faced with a weaker standard sanctioned by DOL. Further, our analysis of commonly-misclassified occupations shows that the independent contractor version of paid work actually has less value for the worker than the employee-status version that the same worker could find – in other words, the worker still bears costs because the independent contractor version of the work likely offers lower pay, fewer benefits, and fewer protections.

In this comment we will estimate these transfers from workers and the social insurance system to employers. The basic structure of this analysis is to take (1) the estimated change in the value of a job to a worker if they are classified as an independent contractor instead of an employee, and (2) the estimated change in payments to social insurance funds if a worker is classified as an independent contractor instead of an employee, and then multiply these figures by the estimated number of workers who will shift to independent contractor status if this rule is finalized. This approach will yield the aggregate impact of the rule on workers and on social insurance system coffers.

In a recent publication, EPI estimated (1) and (2) above for workers in lower-wage, labor intensive occupations most likely affected by the rule, such as call center workers, landscaping workers, janitors and cleaners, home health aides, truck drivers, delivery workers, manicurists, housekeeping cleaners, retail sales workers, security guards, and construction workers.4

The cost to workers in these occupations of being classified as an independent contractor instead of a payroll employee ranges from $6,294 annually for retail sales workers (under extremely conservative assumptions), to $23,266 annually for truck drivers (under less conservative assumptions). Similarly, the annual cost to social insurance funds if a worker is classified as an independent contractor instead of an employee ranges from $600 for manicurists (again under extremely conservative assumptions), to $3,046 for construction workers (again under less conservative assumptions).

Given that we do not have a way to determine where the average impact for those affected by the proposed rule falls in those broad ranges, we simply take the lower bound in both cases, to be extremely conservative. Thus, we assume that the cost to workers is $6,294 annually, and the cost to social insurance programs is $600 per reclassified worker annually.

It should be noted that these lower-bound estimates assume that workers classified as independent contractors are paid not just the full regular pay of a W-2 employee, but also are fully compensated for the value of health insurance and retirement benefits. This is, however, highly unlikely in these occupations. The theory that businesses will not be able to pay less in total compensation to workers if their status shifts from employee to independent contractor—that their base pay will rise to make up for a reduction in benefits—is based on the assumption of perfectly competitive labor markets. There is broad and growing evidence that perfect competition is rare, and that most labor markets do not function competitively—particularly low-wage labor markets like those under consideration here, where workers are more likely to lack the power to bargain for higher wages to compensate for their loss of benefits and increase in taxes when they become independent contractors.5 Further, very low-wage employees whose wage is elevated by the minimum wage could easily see their wage drop when, as independent contractors, they no longer legally must be paid the minimum wage.

How will the share of the workforce who are payroll employees and the share of the workforce that are independent contractors change as a result of this rule?

To begin to answer that question, we need to know how many independent contractors there currently are. There is a great deal of uncertainty around this number (the Department notes that “there are a variety of estimates of the number of independent contractors, and these span a wide range based on methodologies and how the population is defined”).6 The July 2023 Contingent Worker Supplement finds that there were 11.9 million workers who are independent contractors in their main job. This number, however, drastically underestimates the total number of independent contractors by not including workers who do independent contracting on the side, in addition to a payroll job. The Department makes a correction for this issue and estimates that there are 24.8 million individuals working as contractors at a given time. For the sake of the calculations in this comment, we will limit the analysis to the 11.9 million workers the CWS finds are independent contractors in their main job, since workers who do independent contracting as a side job likely work fewer hours and therefore may lose less than the $6,294 we are conservatively assuming workers whose status changes as a result of this rule lose annually. It should be noted that this means we are leaving out many millions of independent contractors and our estimates will, as a result, be extremely conservative for this reason as well. 

How much will independent contracting increase as a result of this rule?

The Department’s proposal would potentially allow companies to legally argue that workers who are now misclassified as independent contractors, or who are working “off the books,” would be legitimately classified as independent contractors under the narrow terms of the proposal. As such, one approach would be to use the percentage of workers misclassified or working off the books under current law to estimate the number of workers who could be reclassified as independent contractors under the proposed rule. However, due to severe data constraints, estimates of the share of workers who are misclassified as independent contractors or working off the books are limited. A 2020 paper estimates that between 12.4% and 20.5% of workers in the construction industry are either misclassified as independent contractors or working off the books.7 Conservatively assuming that the bottom of this range applies more broadly to the lowest-paid quartile of the U.S. labor market, that is 5.1 million low wage workers who may be affected by this rule.8 Of course, these are workers who are already not getting the benefit of being a payroll employee, so the economic impacts described above would not apply. However, this exercise does provide a broad sense of the potential scope of workers affected. Further, even these workers lose something of value under this rule given the current enforcement regime, namely the legal right to the wages and benefits they would receive if they were properly classified. We do not attempt to quantify this effect.

To be exceedingly conservative, we will simply assume that there will be an increase as a result of this rule of 5% in the number of workers who are independent contractors in their main job.9 This translates into an increase of just 595,000 workers who are independent contractors at their main job, given the conservative CWS estimate of 11.9 million workers who are independent contractors in their main job. Multiplying that by our conservative estimate that these workers would lose $6,294 per year yields an aggregate loss to workers of over $3.7 billion annually. Further, social insurance funds would lose at least $357 million annually (595,000 times $600) in the form of reduced employer contributions, meaning this rule also results in a transfer of at least $357 million annually from social insurance funds to employers.10

The NPRM would also have ripple effects in lost benefits and protections that employees are entitled to under other statues. The proposed rule would also extend the weakened definition of employee status to the Family and Medical Leave Act (FMLA) and the Migrant and Seasonal Worker Protection Act (MSPA). Farm workers are already among the most vulnerable, low-paid workers in the U.S., and often face challenges at worksites including poor workplace safety conditions. If farm employers and farm labor contractors have the ability to offload more of their basic responsibilities under MSPA, more farm workers will be at risk of classification as independent contractors and lose even their basic rights under MSPA11, such as to be paid on time or have their working conditions disclosed. More workers would also be at risk of losing access to the right to take job-protected, unpaid family and medical leave under FMLA, which also references the definition of “employee” under the FLSA to determine eligibility for FMLA coverage. The National Partnership for Women and Families has estimated that 15 million workers took advantage of FMLA leave in 2025 alone.12 Protections for break time for nursing mothers—recently expanded under the PUMP Act—are also tied to FLSA employee status. Losing the right to take job-protected time off for illness or the birth of a child, the right to take a break to pump milk, the right to know when you will be paid and to be paid on time—these all specifically conflict with DOL’s stated interest in improving flexibility and satisfaction for workers. This is false flexibility.

The reality of flexible work

The Department focuses on “flexibility and satisfaction” as important non-pecuniary attributes that workers may trade income to receive. However, it is difficult to imagine that there are a meaningful number of workers who would get more satisfaction from doing the same job for substantially less compensation as an independent contractor than for substantially more compensation as a payroll employee. Many workers indeed may value flexibility, but notably, employers are able to provide a huge amount of flexibility to payroll employees if they choose to; the “inherent” tradeoff between flexibility and payroll employment is greatly exaggerated. Workers also highly value other factors, like income stability, which are much less prevalent among independent contractors and are not taken into account here.

In 2024, EPI published a report reviewing the available research and survey data on worker preferences regarding flexibility, stability, and predictability.13 While workers do often prefer flexibility and control over their own schedules, they also want stable, full-time work with predictable pay and benefits.

Employers often incorrectly claim that the FLSA prevents flexible scheduling, but employers control scheduling decisions and can organize work schedules to meet FLSA’s requirements. Employers have long been able to provide flexible schedules and comply with wage and hour laws, and flexible schedules have been negotiated by employers and unions in compliance with the law. Scheduling decisions are the employer’s prerogative (in negotiation with their workers’ union, if there is one), and they can and do set and change schedules in accordance with production demands. Independent contractor status is hardly needed for employers to provide their workers with flexibility.

In conclusion, we urge DOL to withdraw this rule as proposed. The Department should not be in the business of weakening labor protections standards, and should instead seek to vigorously enforce laws against misclassification.

Sincerely,

Samantha Sanders
Director of Government Affairs & Advocacy
Economic Policy Institute

Heidi Shierholz, Ph.D.
President
Economic Policy Institute

Valerie Wilson, Ph.D.
Director, Program on Race, Ethnicity, and the Economy
Economic Policy Institute

Endnotes

1. Ismael Cid-Martinez et al., Misclassifying workers as independent contractors is costly for workers and social insurance systems, Economic Policy Institute, April 2026.

2. Maeve P. Carey, Cost-Benefit and Other Analysis Requirements in the Rulemaking Process, Congressional Research Service, December 9, 2014.

3. 91 Fed. Reg. 9967.

4. Ismael Cid-Martinez et al., Misclassifying workers as independent contractors is costly for workers and social insurance systems, Economic Policy Institute, April 2026.

5. Alan Manning Monopsony in Motion: Imperfect Competition in Labor Markets (Princeton, NJ: Princeton University Press, 2003); Anna Sokolova and Todd Sorensen, Monopsony in Labor Markets: A Meta-Analysis, Washington Center for Equitable Growth, February 2020; Arindrajit Dube, Jeff Jacobs, Suresh Naidu, and Siddharth Suri, “Monopsony in Online Labor Markets,” American Economic Review: Insights 2, no. 1 (March 2020): 33-46, https://www.aeaweb.org/articles?id=10.1257/aeri.20180150.

6. 91 Fed. Reg. 9962. 

7. Russell Ormiston, Dale Belman, and Mark Erlich, An Empirical Methodology to Estimate the Incidence and Costs of Payroll Fraud in the Construction Industry, Institute for Construction Economics Research, January 2020. 

8. Data from the Current Population Survey from the Bureau of Labor Statistics find that there were 163.0 million workers in the U.S. in the first quarter of 2026; 5.1 million = 163.0 million * .25 * .124. 

9. A 5% increase is a conservative assumption, given that the Department is proposing to amend the five-part economic realities test—which has always been interpreted by the Supreme Court in its totality, not weighing any one factor more than another—in a way that will place undue weight on two factors and then narrows those two factors further, making it more likely that workers will be classified as independent contractors and as a result likely leading to a substantial increase in the number of independent contractors. 

10. Some might argue that social insurance funds wouldn’t be hurt by not having employers pay into unemployment insurance and workers’ compensation because independent contractors aren’t eligible for those benefits. However, low-paid independent contractors who lose their contracts and are without work, or get hurt on the job, will be likely to need to depend on safety net programs to survive, so the social insurance system as a whole would still be depleted. 

11. Wage & Hour Division, U.S. Department of Labor, “Fact Sheet #35: Joint Employment and Independent Contractors Under the Migrant and Seasonal Agricultural Worker Protection Act,” revised January 2020. 

12. National Partnership for Women & Families. 2026. Key Facts: The Family and Medical Leave Act (fact sheet), January 2026.

13. Margaret Poydock, Lynn Rhinehart, and Celine McNicholas, Flexible Work: What Workers, Especially Low-Wage Workers, Really Want And How Best To Provide It, Economic Policy Institute, July 2024.