Today, 40 organizations led by the Economic Policy Institute, We Build Progress, the AFL-CIO Tech Institute, and Workshop delivered the letter below urging Congress to center workers in federal AI legislation. 

Dear Member of Congress: 

Employers’ increasing use of AI systems has the potential to affect the lives and livelihoods of workers across the country. Without appropriate guardrails, employers’ integration of these technologies may jeopardize workers’ rights, put workers at risk of discrimination, violate privacy rights, and dramatically impact the economic stability of working families.

These risks posed by technological change are not new. For years, employers have used algorithmic or automated systems and similar technologies in ways that harm workers. Now, the pervasive and growing integration of AI into the workplace is amplifying these risks. These impacts on workers are further exacerbated by persistent power imbalances in the labor market that favor employers. 

It is urgent that Congress take action. It has been nearly two years since the Bipartisan Senate AI Working Group released its roadmap for AI policy, but the Senate has yet to consider comprehensive legislation. AI adoption is moving forward at breakneck speed, and America’s workers cannot afford to wait. 

We applaud members of Congress who have introduced worker-focused legislation addressing issues like civil rights, surveillance in the workplace, and improvements to labor market data. Efforts at broader federal reform must also center the impacts of AI on workers. Under these circumstances, we urge the newly formed House Democratic Commission on AI and the Innovation Economy to center the recommendations of members with expertise in workers’ need for strong labor protections and AI’s impact on the economy. 

The urgency of this moment is further compounded by the Trump administration’s decision to prioritize corporate capture over the public good. In December, after Congress again declined to preempt critical state efforts to regulate AI, President Trump issued an Executive Order that purports to block states from protecting their own residents—a move that blatantly infringes on states’ rights while offering no federal alternative. The administration has doubled down with a national AI legislative framework that would severely curtail states’ ability to regulate AI. Rather than respecting states’ authority to protect their own residents, the administration is doing the bidding of tech oligarchs. 

The AI industry, venture capitalists, and lobbyists spent hundreds of millions of dollars last year pressuring Congress to pass legislation that would prevent state lawmaking. These attempts have failed multiple times because a significant number of members across both parties recognize the dangers posed by AI, while industry actors continue to push for deregulation. 

This is not what the public wants. Recent polling shows a bipartisan consensus in support of AI safety measures: 88% of Democrats and 79% of Republicans favor maintaining existing rules for AI security. Many people want more guardrails on AI: Majorities of both parties are in favor of new regulations to protect society, including 63% of Democrats and 59% of Republicans. 

Federal action is necessary, but it must also leave states room to innovate. Not all states are taking action, so Congress must provide a baseline of protection for people across the country, with a core focus on workers’ rights and livelihoods. 

But federal legislation should be a floor, not a ceiling. Locking the U.S. into a static, insufficient federal framework would guarantee that protections will swiftly become obsolete. It’s important that policymakers do not build a framework that is so narrow or rigid that it fails to keep up with constantly changing AI risks and shifting economic conditions, leaving workers vulnerable to new risks from new tools and practices.

A strong federal framework can create a reinforced system of guardrails to help working people navigate the growing use of AI. Congress has a responsibility to act now—the well-being of our workers and communities depends on it.

Sincerely,

AFL-CIO 

AFL-CIO Tech Institute 

AFT 

American Federation of State, County and Municipal Employees

Americans for Responsible Innovation 

California Initiative for Technology and Democracy

California School Employees Association 

Care in Action

Center for Democracy & Technology 

Center for Oil & Gas Organizing

The Century Foundation 

Communications Workers of America (CWA) 

Consumer Federation of America 

Data & Society 

Economic Policy Institute 

Encode AI 

Future of Life Institute

Interfaith Center on Corporate Responsibility

Jobs With Justice 

The Leadership Conference on Civil and Human Rights 

Legal Aid Justice Center 

Louisiana Progress 

National Action Network 

National Association of Voice Actors 

National Black Worker Center 

National Domestic Workers Alliance 

National Employment Law Project 

National Employment Lawyers Association 

National Institute for Workers’ Rights 

National Partnership for Women & Families 

National Women’s Law Center 

Open MIC (Open Media and Information Companies Initiative) 

Oxfam America

Public Citizen 

Service Employees International Union (SEIU) 

TechTonic Justice 

United Church of Christ Media Justice Ministry

United Food and Commercial Workers International Union 

We Build Progress 

Working Partnerships USA 

Workshop 

Writers Guild of America West

Summary

Roughly 30,000 farmworkers in Colorado, including 4,400 migrant workers recruited by Colorado employers through the H-2A visa program, are treated unfairly under federal and state law. While Colorado took an important step when the state’s overtime law was reformed to make farmworkers eligible—acknowledging the racist policy enshrined in the federal Fair Labor Standards Act that excluded farmworkers from overtime pay—the law nevertheless continues to treat farmworkers unfairly with limited overtime protections compared to those provided to other workers in Colorado.

Farmworkers are some of the lowest-paid employees in the entire U.S. labor market and suffer from high rates of occupational injuries and death. As discussed in this commentary, growth in farmworkers’ very low wages has tracked very similarly to wage growth of other low-wage workers in recent decades. Yet farmworkers must work dramatically more hours than workers out side of agriculture before they can receive any premium for working long hours. There is no economic justification for this unequal treatment. Further evidence for this is the fact that according to the U.S. Department of Agriculture (USDA), labor costs as a share of farm income have not increased in two decades.1

This commentary explains and shows that:

• Overtime pay for farmworkers increases productivity and protects employees from being overworked.
• Farmworkers nationwide earned some of the lowest wage rates in the entire U.S. labor market and only three-fifths of what non-farm workers earn.
• Farmworkers in Colorado earn very low wages—half the average wage earned by all workers in Colorado.
• Wage growth over the past 20 years for farmworkers nationwide has been almost identical to wage growth for other low-wage workers outside of agriculture.
• Real wage growth for farmworkers in Colorado has averaged only 1.5% per year between 2010 and 2024.
• The number of Colorado farms has increased significantly over the past 15 years, suggesting a successful and growing industry in the state.
• California overtime pay standards cover more farmworkers than in Colorado, and outcomes there have shown that providing farmworkers with overtime pay protections on par with those applicable to workers outside of agriculture can be achieved without negatively impacting the farm industry.

If SB 26-121 becomes law, the resulting overtime threshold would further degrade standards for some of the lowest-paid and most vulnerable workers in the U.S. labor market, without improving productivity or benefitting the state’s economy. Farmworkers deserve better: they deserve equal rights and equal pay. While there is some cost to paying workers overtime, keeping the threshold where it is or taking it to 40 hours per week will be partially offset with productivity gains and will benefit farmworkers—most of whom are not employed year-round—by relieving some of the pressure they feel to work as many additional hours as possible, to the detriment of their health, safety, and family life—and possibly to the quality of the nation’s food supply.

Introduction

Farm labor is hard work that sometimes requires very long hours. When it does, workers deserve to be paid fairly for their time. The reason a 40-hour overtime threshold for farmworkers is not already the law should make us wince: When the federal law that governs overtime pay was written in the 1930s, it excluded two job categories that were overwhelmingly held by African Americans—farm laborers and domestic workers. States now have an opportunity to right this historic wrong and level the playing field for all workers

Overtime pay for farmworkers increases productivity and protects employees from being overworked

How do farm owners accommodate paying higher weekly wages when they ask their employees to work overtime? There is of course, some expense associated with overtime pay for farmworkers. But it’s not a dollar-for-dollar cost, so the impact is ultimately modest. The reason is increased productivity.

As we have seen in many other instances, when employers are required to pay higher wages, they make a bigger effort to increase the efficiency of the workplace. We’ve seen this when the minimum wage has been increased. We’ve seen it in unionized businesses. And we’ve seen it already on farms in states like New York and California, when farm owners were required to pay overtime.

What does an increase in productivity on farms look like?

Farm owners may invest in equipment that makes work easier and faster for workers. They may also find ways to organize work that is more effective. Paying overtime provides a real incentive for that. And, overtime pay will reduce the cost of recruitment and training, because it will reduce turnover. That’s something farm operators should value since most claim there are too few farmworkers available to fill open positions.

But importantly, it will also ensure that farmworkers do not work excessive hours just to make enough to survive. Farmworkers in Colorado earn wages that are not much above the state minimum wage, and because of the seasonality of farm work, most are not even able to earn those low wages year-round, reducing their earnings even further. Since farmworkers are not able to earn a living wage year-round, they will feel pressure to work additional hours, to the detriment of their health, safety, and family life—and possibly to the quality of the nation’s food supply.

Colorado’s current overtime threshold is also very different than the one in a state like California. In California, farmworkers earn time-and-a-half overtime after 8 hours in a day or 40 hours in a week.2 They also earn double their regular rate of pay after working 12 hours in a day. Colorado’s overtime threshold has no daily limit, only a weekly one. Even with the additional coverage of overtime for California farmworkers, the number of farm establishments has held steady in the state: going from 16,408 in 2015, the year before the California overtime law was passed, to 16,416 in 2024—suggesting that farms have not been negatively impacted and are still able to operate successfully in the nation’s largest agricultural state.3 

Farmworkers earned some of the lowest wage rates in the entire U.S. labor market in 2024 and only three-fifths of what non-farm workers earn

It is important to discuss and contextualize the wages of the 2.2 million farmworkers in the United States,4 roughly 350,000 of whom are crop farmworkers employed through the H-2A visa program.5 DOL’s National Agricultural Workers Survey (NAWS) shows that two-thirds of non-H-2A crop farmworkers are foreign-born, and that one-third are U.S.-born citizens.6

The agricultural industry has made numerous claims about skyrocketing and unsustainable wage growth for farmworkers, and the industry has lobbied at the state and federal level, pushing for federal actions by the executive branch and legislation to artificially restrain wage growth in the industry. As this letter discusses, many of the major claims made by the industry are not supported by the available evidence.

The most reliable data on farmworker earnings comes from the U.S. Department of Agriculture’s (USDA) National Agricultural Statistics Service (NASS), which conducts the Farm Labor Survey (FLS), the results of which were, until recently, published twice a year in USDA’s Farm Labor report series, with data reported for reference weeks in January, April, July, and October.7 On August 28, 2025, USDA announced that it would discontinue its data collection program and reports, including the FLS,8 thus making 2024 the final full year for which FLS data are available. Before October 2025, FLS data was used by the U.S. Department of Labor (DOL) to set the Adverse Effect Wage Rate (AEWR) for most migrant farmworkers hired in the H-2A program. DOL based the AEWR on the average hourly earnings of nonsupervisory field and livestock workers, as reported by farm operators and by region in the FLS. DOL used the FLS data to set H-2A wages so they reflect current real-world trends in the farm labor market.

The FLS data up to 2024 data show that while there have been some documented real increases over the past three decades, they have not been unreasonably large increases, and they have occurred in a broader context where the wages of farmworkers are extremely low by any measure, even when compared with the hourly earnings of comparable non-farm workers, as well as when compared with average wages for all workers in the United States, and workers with the lowest levels of education (see Figure A).

Figure A

In 2024, the average earnings of all nonsupervisory farmworkers (i.e., combined field and livestock workers in the FLS) was $18.12 per hour. The average farmworker hourly wage in 2024 was just half (52%) of the average hourly wage for all workers in the United States in 2024, which was $34.27 per hour.9 The average farmworker hourly wage in Colorado was less than the national farmworker average, at just $17.84 per hour.

The average hourly wage for production and nonsupervisory non-farm workers—the most appropriate cohort of nonagricultural workers to compare with farmworkers—was $27.56, according to the Current Employment Statistics from the Bureau of Labor Statistics (BLS). In other words, farmworkers earning the national average earned just under 60% of what production and nonsupervisory workers outside of agriculture earned, or three-fifths. In 2024, the farmworker wage gap remained substantial and virtually unchanged from the previous three years. USDA’s ERS shows that between 1990 and 2023, the gap slowly narrowed from 50% to 60% and has described the wage gap between farmworker and nonfarm worker wages as “still substantial, but it is slowly shrinking.”10 

Farmworkers have very low levels of educational attainment and their wages are comparable to workers in other industries with similar educational attainment. According to the NAWS, 27% completed the 10th, 11th, or 12th grade, and only 16% completed some education beyond high school.11 Farmworkers earn the same or less than the two groups of non-farm workers with the lowest levels of education in the United States: Farmworkers earned 10 cents an hour more than the average wage earned by workers without a high school diploma ($18.02), but earned $5.61 less per hour than the average wage earned by workers with only a high school diploma ($23.73). Farmworkers in Colorado earned less than workers without a high school diploma ($17.84 vs $18.02).

Farmworkers in Colorado earn very low wages—half the average wage earned by all workers in Colorado

There are roughly 30,000 farmworkers in Colorado, including 4,400 migrant workers recruited by Colorado employers through the H-2A visa program. As noted above, in 2024, at the state level in Colorado, USDA’s FLS shows that the average hourly wage for farmworkers in Colorado (the combined average wage for field and livestock workers) was $17.84. Figure A also shows that the average wage for all workers in Colorado in 2024 was $33.63, according to the Occupational Employment and Wage Statistics (OEWS) of the Bureau of Labor Statistics.12 In other words, farmworkers just earned 53%, roughly half, of the average wage that all Colorado workers earned.13 And as noted above, most farmworkers are not employed year-round. Despite these extremely low wages, farmworkers in Colorado work in some of the most difficult and dangerous conditions while providing an essential function for the economy and state. Thus they deserve more protections under the overtime law, not fewer.

Industry claims about the increases in farmworker wages ignore the fact that wage growth over the past 20 years has been almost identical to wage growth for other low wage workers

The value and the rate of increase of the Adverse Effect Wage Rage (AEWR) for H-2A farmworkers has become a hot-button issue and many claims about its impact have been made over the years by representatives of industry. These are relevant to examine because the AEWR wages up until 2025 represented the wages that farm operators reported they were paying to their farmworkers in response to the USDA’s Farm Labor Survey. Thus, they represent the best data available on average farm wages at the national and regional level.

Many of the claims about wage growth for farmworkers made by industry advocates and even the U.S. Department of Labor (DOL) about year-to-year increases often do not adjust for inflation, which overstates the actual increase in terms of its dollar value. This is a basic mistake that misleads—and it misleads particularly during times of relatively rapid inflation, like the post-pandemic period. DOL echoes these misleading claims from industry advocates and makes their own false claim in the preamble to the October 2025 AEWR Interim Final Rule, making the year-over-year increases in farmworker wages seem greater than they truly are. DOL notes that the national average AEWR—i.e., the average combined field and livestock worker wage reported by farm operators nationwide—has more than doubled in nominal terms over 20 years from $8.56 in 2005 to $17.74 in 2025.14 But DOL’s own CPI Inflation Calculator adjusts the value of $8.56 in September 2005 to $13.99 in September 2025, resulting in a real increase of just over one-quarter over two decades, at 26.8%, which over that period averages out to just 1.2% per year.

If we examine the same period for other low-wage workers in nominal terms, we also see that wage growth for farmworkers as represented by the AEWR, is in line with—nearly identical to—nominal wage growth for other low wage workers in the United States. Figure B below shows annualized wage growth for workers paid at the 20^th percentile wage, as well as the median wage for workers with less than a high school education—both of which are good measures for typical low-wage workers. Both saw annual nominal wage growth that was at 3.5% between 2005-2025, the period that DOL identifies. Farmworkers earning the national average farmworker wage—over that same period saw annualized wage growth of 3.7%, nearly identical to other typical low-wage workers. Thus, DOL’s main example in its H-2A wage regulation of runaway wage growth for farmworkers does not hold water.

Figure B

Real wage growth for farmworkers in Colorado has averaged only 1.5% per year between 2010 and 2024

While Figure B looked at nominal wage growth over the past 20 years, EPI has previously calculated the total real wage growth for farmworkers (i.e., after adjusting for inflation) in every state between 2010 and 2024.15 We found that in Colorado, the average farmworker wage in 2010 was $14.51 (adjusted to 2024 dollars), growing to $17.84 fifteen years later in 2024. This amounts to a total increase of $3.33 over 15 years (in 2024 dollars), or 22.9%. Farmworkers in Colorado averaged a real wage increase of 1.5% per year over the 15-year period.

The number of Colorado farms has increased significantly over the past 15 years, suggesting a successful agricultural industry in the state

One common argument from farm operators is that if the wages of farmworkers are too high, those high wages will put them out of business. But according to the BLS’s Quarterly Census of Employment and Wages (QCEW), the number of agricultural establishments in Colorado has increased significantly over the past 15 years. QCEW data show that the number of agricultural establishments in Colorado averaged 1,412 between 2010 and 2012. By 2024, the number of agricultural establishments had increased to 1,812, an increase of 28.3%. The 2022-2024 average number of agricultural establishments was 1,856.16

Agribusiness representatives may claim that agricultural establishments in Colorado will be forced to close or will decide to move their operations to other U.S. states because of higher labor costs associated with farmworkers being entitled to overtime pay, but the reality is that the number of agricultural establishments has been increasing steadily, even as farmworker wages have risen modestly, suggesting that both farm owners and farmworkers can mutually benefit from a growing industry.

Conclusion: The Colorado legislature should not further degrade standards on farms by expanding the 56-hour overtime threshold—and should instead provide farmworkers with equal rights in the workplace by providing them overtime after 40 hours

The annual average real wage growth of 1.5% per year over 15 years represents moderate wage growth for farmworkers and suggests a relatively tight labor market for farmworkers. However, it represents little improvement in job quality for workers that have been exempted from key labor laws and wage and hour standards, who frequently toil for long hours in difficult conditions without any pay premium, and who consequently still earn only 50% to 60% of the wage earned by comparable nonsupervisory workers outside of agriculture (see Figure A and discussion above). It would take many more years of comparatively faster wage growth for farmworkers to begin to approach even three-fourths of what nonsupervisory workers earn outside of agriculture.

Further degrading standards for some of the lowest-paid and most vulnerable workers in the U.S. labor market will not improve productivity or benefit the state’s economy; it will do the opposite, taking money out the pockets of workers who live paycheck to paycheck and spend those earnings on necessary goods and services. All while making a minimal impact on the overall share of farm income that farm operators spend on workers’ wages.

Instead of passing SB 26-121, the state legislature should set a reasonable minimum standard for the wages paid to farmworkers, and that standard should be no different than the standard set for most other workers in Colorado, which is a 40-hour overtime threshold.

Endnotes 

1. Economic Research Service, “Labor Cost Share of Total Gross Revenues,” in “Farm Labor,” U.S. Department of Agriculture, Updated November 18, 2025.

2. See Department of Industrial Relations, “Overtime for Agricultural Workers,” State of California, last updated October 2023.

3. See Quarterly Census of Employment and Wages, Bureau of Labor Statistics, Series Id: ENU5100020511, Series Title: Number of Establishments in Private NAICS 11 Agriculture, forestry, fishing and hunting, for all establishment sizes in California – Statewide; Owner: Private, for 2010-24.

4. As counted by the latest Census of Agriculture from the U.S. Department of Agriculture, 2022.

5. See Daniel Costa and Ben Zipperer, “Trump’s new H-2A wage rule will radically cut the wages of all farmworkers: New estimates show farmworkers stand to lose $4.4 to $5.4 billion annually under DOL’s updated Adverse Effect Wage Rate,” Working Economics blog (Economic Policy Institute), November 26, 2025.

6. Wenson Fung, Kimberly Prado, Amanda Gold, Andrew Padovani, Daniel Carroll, and Emily Finchum-Mason, Findings from the National Agricultural Workers Survey (NAWS) 2021–2022: A Demographic and Employment Profile of United States Crop Workers, Research Report no. 17, JBS International for the Employment and Training Administration, U.S. Department of Labor. September 2023.

7. See National Agricultural Statistics Service, “Agricultural (Farm) Labor,” for more background and to access Farm Labor Reports, U.S. Department of Agriculture.

8. Federal Policy Watch, “USDA ends the Agricultural (Farm) Labor Survey, the U.S.’s only survey of agricultural employers,” Economic Policy Institute, September 3, 2025.

9. Economic Policy Institute, State of Working America Data Library, “Hourly wage, average – Average real hourly wage (2024$),” 2025.

10. Economic Research Service, “Wages of Hired Farmworkers” in “Farm Labor,” U.S. Department of Agriculture, Updated November 18, 2025.

11. Wenson Fung, Kimberly Prado, Amanda Gold, Andrew Padovani, Daniel Carroll, and Emily Finchum-Mason, Findings from the National Agricultural Workers Survey (NAWS) 2021–2022: A Demographic and Employment Profile of United States Crop Workers, Research Report no. 17, JBS International for the Employment and Training Administration, U.S. Department of Labor. September 2023.

12. https://data.bls.gov/oes/#/area/0800000

13. A note about the data: The wage cited is for Colorado farmworkers in USDA’s FLS represents the wage reported for the Mountain II region, which surveys farm operators in Colorado, Nevada, Utah. USDA’s FLS conducts wage surveys by multistate region, except for California which USDA surveys as its own individual region.

14. Employment and Training Administration, Adverse Effect Wage Rate Methodology for the Temporary Employment of H-2A Nonimmigrants in Non-Range Occupations in the United States, U.S. Department of Labor, Interim Final Rule, 90 Fed. Reg. 47914, at 47923 (October 2, 2025).

15. See Table 1 in Daniel Costa, EPI comment on DOL’s 2025 Interim Final Rule modifying the AEWR methodology for H-2A farmworkers, Economic Policy Institute, December 1, 2025.

16. See Quarterly Census of Employment and Wages, Bureau of Labor Statistics, Series Id: ENU5100020511, Series Title: Number of Establishments in Private NAICS 11 Agriculture, forestry, fishing and hunting, for all establishment sizes in Colorado – Statewide; Owner: Private, for 2010-24.

Further, this move radically undermines what Trump says his own goal is: lowering U.S. interest rates to spur faster economic growth. Instead, it will likely raise interest rates over the long term and lead to higher costs for working people.

Why will this happen? The interest rates that truly influence economic growth are long-term rates generally set in financial markets. These longer-term rates are usually strongly influenced by the Fed’s decisions over the shorter-term rates it controls directly—but that’s only because the Fed has, until now, been seen as a non-political, evidence-based institution. If the Fed instead becomes politicized, its influence will falter, and changes it makes to short-term rates will have far less effect on the long-term rates that influence economic growth.

Presidential capture of the Fed would signal to decision-makers throughout the economy that interest rates will no longer be set on the basis of sound data or economic conditions—but instead on the whims of the president. Confidence that the Fed will respond wisely to future periods of macroeconomic stress—either excess inflation or unemployment—will evaporate. As a result, investors will demand higher premiums to hold on to U.S. Treasury bonds (and other long-term bonds), because without faith that the Federal Reserve will tamp down inflationary pressures when they appear, they will need reassurance—in the form of higher long-term interest rates—to hold on to these investments.

These higher long-term rates will ripple through the economy—making mortgages, auto loans, and credit card payments higher for working people—and require that rates be held higher for longer to tamp down any future outbreak of inflation. In the first hours after Trump’s announcement, all of these worries seemed to be coming to pass.

The source of the allegation of mortgage fraud against Governor Cook is also extremely concerning: a public announcement by the head of the Federal Housing Financing Authority (FHFA), the agency that oversees Fannie Mae and Freddie Mac. The FHFA has access to mortgage information for tens of millions of U.S. households—access that could be abused by a politically-motivated agency looking to harm perceived political opponents of the president. This seems clearly to be what is happening in the Cook case. Under an honest and well-run administration, any potential impropriety identified by FHFA staff in the normal course of their activities would have been referred (without public notice) to the Department of Justice, which would have conducted an investigation without any public comment. Only if the allegations rose to the level of an indictment would a public announcement be made. That the FHFA has been weaponized to find damaging allegations against a perceived political opponent of the president is deeply worrying.

EPI urges the courts to act quickly to overturn this unlawful dismissal and to reaffirm the independence of the Federal Reserve, which is critical to the health of our economy and our democracy.

BLS is one of the most respected statistical agencies in the world, known for its methodological rigor, independence, and transparency. It has long produced reliable economic data that are a crucial input to economic decisions across the country. The economy runs on reliable data. Businesses use these numbers to decide whether to hire or expand. The Federal Reserve uses them to decide when and how much to change interest rates. State and local governments use them to plan budgets. Trump’s attempt to politicize BLS means that policymakers and the public wouldn’t be able to trust the data. If this happens, confidence in U.S. data will collapse and reasonable economic decision-making will be impossible. This manufactured chaos will reduce business investment and consumer spending, making a recession—and soaring unemployment—far more likely in coming months. Between illegal firings of public servants, starving data agencies of needed resources, and now political intimidation, the U.S. looks set to run into the next economic downturn flying blind. The cost of this incompetence will be felt by working people first.

BLS is one of the most respected statistical agencies in the world, known for its methodological rigor, independence, and transparency. The president’s belief that the BLS commissioner personally “produced” the jobs numbers is preposterous and shows a complete misunderstanding of how government statistical agencies operate. These data are the product of careful work by hundreds of expert economists, statisticians, and civil servants following transparent, well-established methodologies.

Dr. Erika McEntarfer is a highly respected labor economist with deep expertise in labor market data. Like all BLS commissioners, she is bound by strict norms of nonpartisanship and statistical integrity. To fire her over an official data release—simply because the numbers do not serve a particular political narrative—is a deeply dangerous attack on the foundations of a functioning democracy. 

And it’s not just undemocratic—it’s economically dangerous. The economy runs on reliable data. Businesses use these numbers to decide whether to hire or expand. The Federal Reserve uses them to set interest rates. State and local governments use them to plan budgets. If policymakers and the public can’t trust the data—or suspect the data are being manipulated—confidence collapses and reasonable economic decision-making becomes impossible. It’s like trying to drive a car blindfolded. This manufactured chaos will reduce business investment and consumer spending, making a recession—and soaring unemployment—far more likely in coming months.

The firing of the BLS commissioner follows a week of reporting that almost 20% of price inflation data was unable to be collected for the month of July, stemming from severe resource constraints. Between illegal firings, starving data agencies of needed resources, and now political intimidation, the U.S. looks set to run into the next economic downturn flying blind. The cost of this incompetence will be felt by working people first.

The tax breaks for the rich are so huge—$100,000+ per year for the richest 0.1%—that even after gutting aid for the most vulnerable, this bill will still increase the national debt by nearly $4 trillion. To cover up how grotesque this is, Republicans have invented a new, completely bogus way to measure the bill’s costs.

The bill also turbocharges an authoritarian-style immigration regime—funding internment camps, mass surveillance, and waves of deportations that will kill millions of jobs.

And, surprise, the GOP structured the bill’s provisions along deeply cynical political timelines. The rich get their tax cuts before the midterms. Some of the most painful (and unpopular) cuts to families? Not until after. It is shameless. And that means that—by design—this bill will cause economic pain that will unfold slowly but steadily over many years. It’s engineered to dodge accountability.

But make no mistake—the pain will be devastating. Kids will lose food assistance. Families will lose health care. The economic shock will hit hardest in communities least equipped to withstand it.

It’s a perfect storm of long-term economic sabotage. This bill will weaken economic growth over the next decade, making this country poorer. All to give huge tax cuts to the richest.

The most acute and immediate damage stemming from this bill would be felt by the millions of American families losing key safety net protections like Medicaid and Supplemental Nutrition Assistance Program (SNAP) benefits. The Medicaid cuts constitute a sad step backward in the nation’s commitment to providing access to health care for all. Proponents of the House bill often claim that these Medicaid cuts can be achieved simply by imposing work reporting requirements on healthy, working-age adults. But healthy, working-age adults are by definition not heavy consumers of health spending, so achieving the budgeted Medicaid cuts will obviously harm others as well.

Medicaid provides health insurance coverage for low-income Americans, but this includes paying out-of-pocket health costs for low-income retired Medicare recipients and providing nursing home and in-home care services for elderly Americans. Medicaid also covers 41% of all births in the United States, including over 50% of all births in Louisiana, Mississippi, New Mexico, and Oklahoma. Work reporting requirements will obviously yield no savings from these Medicaid functions.

Besides providing affordable health care to families, Medicaid is also crucial to state budgets and hospital systems throughout the country—particularly in rural areas. In 2023, the federal government sent $615 billion to state governments to cover Medicaid spending; this federal contribution accounted for over 75% of total state Medicaid spending in more than 19 states. Rural hospitals in states that accepted the Medicaid expansion that was part of the Affordable Care Act were 62% less likely to close than rural hospitals in non-expansion states.

In addition to Medicaid, the House bill also significantly cuts SNAP. These steep cuts to the social safety net are being undertaken to defray the staggering cost of the tax cuts included in the House bill, including the hidden cost of preserving the large corporate income tax cut passed in the 2017 tax law. But even these sharp spending cuts will pay for far less than half of the tax cuts (not even including the cost of maintaining the corporate income tax cuts of the 2017 law).

U.S. structural deficits are already too high, with real debt service payments approaching their historic highs in the past year. The House bill layers $3.8 trillion in additional tax cuts ($5.3 trillion if all provisions are made permanent) on top of these existing fiscal gaps—and these tax cuts are overwhelmingly tilted toward the highest-income households. Even with the safety net cuts, the House bill leads to public debt rising by over $3 trillion in coming years (and over $5 trillion over the next decade if provisions are made permanent rather than phasing out). The higher debt and deficits will put noticeable upward pressure on both inflation and interest rates in coming years.

The combination of cuts to key safety net programs like Medicaid and SNAP and tax cuts disproportionately benefiting higher-income households means that the House budget constitutes an extremely large upward redistribution of income. Given how much this bill adds to the U.S. debt, it is shocking that it still imposes absolute losses on the bottom 40% of U.S households (if some of the fiscal cost is absorbed in future bills with extremely high and broad tariffs, the share of households seeing absolute losses will increase rapidly).

The United States has a number of pressing economic challenges to address, many of which require a greater level of state capacity to navigate—capacity that will be eroded by large tax cuts. The House bill addresses none of the nation’s key economic challenges usefully and exacerbates many of them. The Senate should refuse to pass this bill and start over from scratch on the budget.

Daron Acemoglu
MIT Economics

Peter Diamond
MIT Economics

Oliver Hart
Harvard University

Simon Johnson
MIT Sloan School of Management

Paul Krugman
Graduate Center, City University of New York

Joseph Stiglitz
Columbia University

In direct contradiction to promises from President Trump and the reconciliation bill’s backers that they won’t cut Medicaid or cause pain for working-class families, this bill would:

• Kick millions of low-income people off of Medicaid and threaten operations for strained hospital systems across the country, particularly in rural areas;
• Make it harder for parents of children as young as age seven to qualify for food stamps;
• Put millions of kids at risk of poverty by excluding their families from the Child Tax Credit.

In exchange for these massive spending cuts to justify a $3.8 trillion tax cut for the wealthy, this bill offers workers almost nothing—just paltry tax gimmicks that put pennies into the pockets of a small sliver of the working class. The bill’s immigration provisions are purely punitive—imposing fees on migrants making asylum claims and giving the Trump administration billions of dollars to unleash internal immigration enforcement within our borders, looking for any excuse to deport or harass immigrants and their families. Voting for this bill’s passage represents an ultimate betrayal of U.S. workers and the economy. The Senate should reject this bill.

This is a fair-enough characterization of what the blog post said about the steel and aluminum tariffs it was analyzing, but it misleads in applying this conclusion to the tariffs being proposed yesterday (and in previous weeks) by the current Trump administration. Specifically, the steel and aluminum tariffs examined in that EPI blog post were narrowly targeted and applied to roughly $50 billion of imports, less than 2% of the imports that yesterday’s announced tariffs would apply to. In terms of assessing the effect of tariffs, this large difference in scale matters crucially.

It is hard to overstate what a radical change yesterday’s tariff announcements are relative to either the steel and aluminum tariffs we examined in that blog post or even to the full suite of tariffs passed in the first Trump administration. The first Trump administration’s tariffs were far narrower and only undertaken after a lot more study and consultation. For example, seven months into his first term, President Trump ordered the United States Trade Representative to investigate Chinese trade practices. The report was released in March 2018 and found a number of instances of China engaging in unfair practices. In that same month (over a year into the term), the Trump administration announced tariffs on up to $60 billion worth of imports from China.

By the end of the first Trump administration, further tariffs had been implemented. With these additional tariffs, roughly $380 billion worth of goods were subject to these taxes, which had been phased in over four full years.

Now, Trump is imposing tariffs on essentially all goods imports—more than $3 trillion—after just over two months into his term. Tariffs can be a legitimate and useful tool in industrial policy for well-defined strategic goals, but broad-based tariffs that significantly raise the average effective tariff rate in the United States are unwise. Further, the second Trump administration’s rationale, parameters, and timeline for tariffs have been ever-shifting. As the original post cited by the administration argues, tariffs should not be a goal unto themselves, but a strategic tool to pair with other efforts to restore American competitiveness in narrowly targeted industrial sectors.