This is a guest blog from Owen Herrnstadt, former Chief of Staff of the International Association of Machinists and Aerospace Workers and former member of the Board of Directors of the U.S. Export-Import Bank. 

International labor and human rights standards are widely recognized but rarely enforced. The Universal Declaration of Human Rights, the conventions and declarations of the International Labor Organization (ILO), the Organization for Economic Co-operation and Development (OECD) guidelines, and the performance standards of international financial institutions all set out expectations for how workers should be treated, yet each lacks a meaningful mechanism for compelling compliance. Trade law has emerged as a possible exception. Renegotiation of the United States-Mexico-Canada Agreement (USMCA) offers an opportunity to significantly strengthen efforts to effectively enforce labor and human rights in a regional trade agreement. 

While USMCA’s labor chapter and its annexes—including the Rapid Response Labor Mechanism (RRM)—represent the most ambitious labor enforcement architecture in any U.S. trade agreement, serious criticism remains over its impact on improving labor and human rights. Mexico’s reforms have not gone far enough, implementation has raised serious questions, the RRM—while representing a significant step forward—is narrow, slow, and procedurally burdensome, and the wage and trade balance outcomes the agreement was meant to improve have not moved in the right direction.

Principal recommendations

In the USMCA joint review, the United States should:

1. remove the footnote limiting labor rights obligations only to the 1998 ILO Declaration on Principles and Rights at Work;

2. replace the “sustained or recurring” requirement for a labor violation to a significance-based standard;

3. expand the RRM to all core labor standards explicitly referenced in the ILO’s 1998 Declaration on Principles and Rights at Work and simplify its evidentiary and procedural requirements; 

4. require at minimum National Labor Relations Board (NLRB)-style “laboratory conditions” for union elections in Mexico, including prohibiting incumbent unions from administering them;

5. limit the use of the 30% support requirement for initial union elections to prevent misuse by illegitimate (protection) unions.

6. strengthen and extend strict rules of origin and Labor Value Content (LVC) requirements to other sectors, convert LVC “to a simple wage floor,” raise the $16 per hour reference, and tie it to sectors and the rules of origin; and 

7. strengthen traceability and customs enforcement to close third-country transshipment loopholes including the utilization of “substantial transformation.”

The evolution of labor standards in U.S. trade agreements

NAFTA and the side agreement model

When the North American Free Trade Agreement (NAFTA) was negotiated in 1993–94, labor unions and other civil society groups argued that labor standards were as much an economic issue as intellectual property or other traditional trade concerns. The theory was straightforward: If a country such as Mexico could hold down labor costs by suppressing union formation and collective bargaining, or by permitting discrimination, child labor, or forced labor, its goods would unfairly undercut goods made in the United States. U.S. companies and workers would lose market share, wage growth and bargaining power would be chilled, and plant closures would hollow out local economies. 

Despite these arguments, labor standards were left out of the core text of NAFTA. They were relegated to a side agreement—the North American Agreement on Labor Cooperation (NAALC)—outside NAFTA’s principal enforcement mechanisms. The predictable results followed. U.S. employers used the threat of relocation to Mexico to restrain wages, plants closed, and Mexican workers saw no meaningful improvement in wages, union rights, or discrimination. The maquiladora sector became a symbol of the gap between the treaty’s promise and workers’ lived experience.1

From Seattle to Peru

Pressure to include enforceable labor standards grew through the second half of the 1990s. At the 1999 World Trade Organization (WTO) Ministerial Conference in Seattle, hundreds of members of the International Association of Machinists and Aerospace Workers (IAM Union) who depended on exports led the peaceful protests over the absence of labor standards in the WTO framework. 

Labor’s central argument was that artificially suppressing labor costs by failing to recognize and enforce fundamental human rights amounts to an unlawful subsidy and thus an unfair trade practice. President Clinton’s proposal concerning a WTO Working Group on Trade and Labor never gained a consensus and went nowhere at the WTO talks.

The next major step was the U.S.-Peru Trade Promotion Agreement (as revised in 2007), the first U.S. agreement to place labor standards in the core text rather than a side accord. The parties committed to “adopt and maintain” laws consistent with the 1998 Declaration, and all core standard obligations became enforceable through the same dispute settlement used for commercial matters, making trade sanctions formally available for labor violations. 

Despite the improvement, several serious criticisms of this model remained, including: 

1. Reference only to the 1998 Declaration was inadequate. Tying the labor chapter to the Declaration rather than to the Conventions reduced complex standards, such as the right to organize and bargain collectively, to a general principle. Critical questions, such as the controversial use of permanent striker replacements, were left unresolved. A footnote excluding the Conventions, Recommendations, and ILO supervisory-body guidance from consideration furthered muddied possible interpretations of the Declaration.
2. The trade-effect requirement. A violation had to affect trade or investment, a burdensome requirement that most workers would not know and had no practical means of proving.
3. The “sustained or recurring” requirement. A violation had to be sustained or recurring, eliminating isolated acts of retaliation even when the act itself (e.g., an unlawful discharge) could crush the right to form a union.

USMCA’s labor architecture

Chapter 23

USMCA, negotiated during the first Trump administration and in force since July 1, 2020, improved on the Peru model through footnotes, new annexes, and new rules of origin. Among other things, it added a footnote in Chapter 23 creating a rebuttable presumption that a violation relates to trade or investment.  

However, USMCA retained the footnote limiting interpretation of labor standards to the 1998 Declaration and preserved the “sustained or recurring” requirement.

Annex 23-A: Mexico’s labor law reform

Annex 23-A sets out specific requirements for reforming Mexican labor law in order to effectively include, in practice, international labor rights. One of the goals of the reform was to eliminate the presence of “protection unions” controlled by corrupt union leaders, companies, or political parties, which did not represent workers’ interests and which many workers did not even know represented them. The Annex required new Mexican law providing for: 

1. personal, free, and secret elections;
2. ready accessibility of collective bargaining agreements (CBAs) to covered workers prior to votes;
3. votes to “legitimize” existing CBAs;
4. votes on the formation of new unions; and
5. a new, independent system of labor courts.

Annex 31-A: The rapid response labor mechanism

Annex 31-A establishes that, for Mexico, a denial of rights occurs upon a violation of the new Mexican labor law itself rather than of the 1998 Declaration. The violation must occur at a “covered facility,” one producing goods or supplying services traded between the United States and Mexico in priority sectors such as automotive, aerospace, and steel. A Rapid Response Labor Mechanism panel then determines whether a denial of rights has occurred. Remedies range from suspension of preferential tariff treatment to, ultimately, blocking entry of the goods into the United States. 

Assessment: results and deficiencies

Results to date

Despite improvement, USMCA fell far short of providing the effective enforcement of labor and human rights standards that were needed. During its negotiation, congressional testimony warned that these shortcomings would negatively impact the agreement’s effect on workers’ rights:

“First, enforcement must be adequately funded and staffed, independently administered and monitored, and must provide swift relief for victims of labor rights violations and sanctions for states and employers that violate those rights; Second, since enforcement is meaningless if the substantive standards to be enforced are weak, labor obligations must be based on strong clear standards based on internationally recognized rights and reports established by the International Labor Organization, an arm of the United Nations; and, third, labor obligations must cover broad sectors of workers and remove obstacles that narrow the types of labor violations that are recognized and actionable under the agreement.”

Five years into its implementation, the results of USMCA have, predictably, been disappointing. The U.S. trade deficit with Mexico grew from $111.0 billion in 2020 to $197.0 billion in 2025.

The wage gap between U.S. and Mexican manufacturing workers remains substantial; Mexican manufacturing wages remain a fraction of U.S. manufacturing wages. Auto production and other manufacturing jobs did not move back to the U.S. from Mexico. And despite Mexico’s labor law reforms, implementation and enforcement remain serious problems: Many workers still do not know whether they are represented by a union, protection unions remain widespread, and questions persist about the integrity of union elections and CBA ratification votes. The Independent Mexico Labor Expert Board’s (IMLEB) 2021 and 2025 reports documented continuing obstacles in the transition away from Mexico’s protection contract system and identified significant implementation concerns.

Deficiencies in Mexico’s implementation

1. Election integrity. Elections in many cases are not secret and would not meet the “laboratory conditions” required for NLRB elections: Votes are taken by show of hands, government monitors appear only at the moment of the vote, and workers report not understanding what they were voting on. One study found that workers in the maquiladora region were often unaware of what a legitimation vote entailed. 
2. Statutory loopholes exist. Legitimizing an existing, often “protection,” union and CBA requires 50% plus one vote, but forming a new union requires only 30% to negotiate a CBA. Some entrenched unions skipped legitimation and simply registered as “new” unions to clear the lower bar. Of roughly 10,773 initial CBAs registered under the new system, a substantial number appear to reflect this loophole, alongside more than 30,000 legitimation votes conducted to date.
3. Credibility of oversight. Mexican authorities reported over 16,000 verification elections between March and July 2023 alone—roughly 1,000 per week—a volume that raises doubts about whether the government could meaningfully oversee each one. 
4. Penalties have not been imposed. Mexico’s new federal labor center has not imposed penalties on employers, even in RRM cases where both governments agreed that a violation occurred and a remedy was warranted, undermining the mechanism’s deterrent effect.

Deficiencies in the RRM’s design

1. Narrow scope. The RRM covers only Mexico’s labor law reform, not the 1998 ILO Declaration of Principles and Rights in the Workplace and is limited to freedom of association and collective bargaining, not discrimination, child labor, forced labor, or health and safety.
2. High evidentiary burden. The United States Trade Representative (USTR) and the Department of Labor must find “sufficient credible evidence,” often demanding proof that workers have little or no practical means of obtaining.
3. Cumbersome procedure. U.S. rules require petitioners alleging violations across multiple facilities of a single company to file separate complaints for each facility.
4. Burdensome “covered facility” requirement. Mexico’s process places an unreasonable burden on petitioners to show that a workplace produces goods or supplies services traded with the United States or competing in the U.S. market. Workers in general and especially those in a supply chain often lack that information.
5. Serious violations excluded from the “sustained or recurring” standard. It ignores serious isolated violations—such as retaliatory firing or harassment—used to deter organizing, and remains ambiguous in application.
6. Rapid mechanism not rapid. Some RRM matters have remained unresolved for more than a year, while cases proceeding through the first stages of the mechanism have often been resolved in months.

Recommendations for the USMCA joint review

USMCA represents an important step toward providing a framework to enforce labor and human rights. However, if it is to be effective, much more work needs to be done. Some suggestions for improving USMCA include the following: 

1. Eliminate the footnote limiting labor standards to the 1998 Declaration, which perpetuates ambiguity over what freedom of association and collective bargaining require in practice.
2. Replace “sustained or recurring” with a standard based on the significance of the violation.
3. Eliminate the “covered facility” requirement, or presume coverage for any facility in a priority sector in either country.
4. Adopt a uniform “good faith belief in denial of rights” standard for both interagency U.S. claims and Mexican filings. Eliminate any requirement that a U.S.-side violation be established by an enforced NLRB order before a claim may proceed.
5. Expand the RRM to cover all core labor standards, including non-discrimination, child labor, forced labor, and health and safety.
6. Require “laboratory conditions” for union elections comparable with NLRB standards. Bar incumbent unions from any role in administering elections, including selecting neutrals or counting ballots. Strengthen worker awareness and education about their rights in Mexico.
7. Limit the use of the 30% support requirement for initial union elections to prevent misuse by illegitimate (protection) unions. Institute a credible, adequately resourced legitimation process with real oversight capacity.
8. Extend coverage to other sectors such as agriculture.
9. Extend and tighten strict rules of origin from automobiles to all priority manufacturing sectors. Add due-diligence provisions requiring tracing of the true origin of parts and materials, closing the China and other non-party routes. Enforce rules of origin through active government oversight rather than importer self-reporting.
10. Expand Labor Value Content to other manufacturing sectors, convert LVC “to a simple wage floor,” raise the $16 per hour reference, and tie it to sectors and the rules of origin.

Notes

1. See also Harley Shaiken, Owen Herrnstadt, and Catha Worthman, “Crossing Borders: Trade Policy and Transnational Labour Education,” Labour, Capital and Society, vol. 35, no. 2 (November 2002), pp. 342–368.