Corporate Accountability by Treaty: The New North American Rapid Response Labor Mechanism

Introduction

For several decades, civil society has sought to impose greater responsibility on companies for cross-border social wrongs. Multiple legal subfields and initiatives have emerged to take on this work: corporate social responsibility (CSR);¹ business and human rights (BHR);² responsible business conduct codes;³ environment, social, and governance (ESG) standards;⁴ and corporate due diligence schemes,⁵ among them. These many projects have culminated in national legislation across a range of jurisdictions imposing reporting and compliance requirements on companies, as well as in the development of soft law standards.⁶ Despite these efforts, international treaty initiatives to create binding obligations on companies and to require them to adopt certain human rights principles have made limited progress.⁷

Unexpectedly, a novel tool embedded in a 2020 trilateral economic agreement among the United States, Mexico, and Canada has emerged with the potential to make inroads in this space. This mechanism, called the Facility-Specific Rapid Response Labor Mechanism (RRM), seeks to achieve corporate accountability goals in the area of freedom of association and collective bargaining. The RRM is the first of its kind in that it allows one state party to enforce the domestic labor law of another party against a particular worksite in the territory of that second state. Never before have governments agreed to a right of action to regulate worksites on one another's territory in this way for labor rights violations.

By advancing these aims through a trilateral economic agreement and tying them to the agreement's major commercial impacts, the RRM shifts corporate accountability debates to a new institutional home. The RRM has a straightforward and narrow labor rights ambition, but, to some, it holds promise to achieve not just that aim but also those held out by these other legal and social movements.⁸ It could bring together the CSR, BHR, and like-minded communities with the labor movement in new ways. This Current Development introduces this unique international legal instrument and surveys its function in the first three years of its operation.

Progressives throughout North America have lauded the RRM as a major step forward in defending workers’ rights.⁹ Its novelty turns on the fact that it is a corporate responsibility tool with hard sanctions in a binding trade agreement. Unlike other economic agreement mechanisms, the RRM targets companies for their social harms rather than countries for their regulatory failures.

Consider an example: Workers at an auto parts facility in Mexico organize to request the execution of a collective bargaining agreement but are prevented from doing so by the company and an undemocratic union. Without the RRM, the aggrieved workers could file a complaint with the Mexican Ministry of Labor and Social Welfare or pursue a remedy in the Mexican courts for this violation of Mexican labor law.¹⁰ If this were to happen repeatedly, civil society groups might highlight the problem, urging action by local or national government officials. But through the RRM, U.S. officials can issue an order, threatening goods from the facility with high tariffs upon import into the United States, that has the potential to force management to halt work at the factory until the union voting exercise is redone.¹¹ The idea is that without the ability to send its goods into the United States expeditiously and tariff-free, which may cause it to lose its customers throughout the supply chain, the manufacturer will move quickly to address the labor problem.

Due diligence efforts in countries such as France and Germany, or European Union efforts toward the same, appear timid next to this blunt instrument with extensive cross-border reach. And yet, the RRM is far from perfect. In its first three years, the Mechanism has proven powerful in achieving corporate change and worker benefits in a few targeted instances and perhaps even in complementing Mexico's own efforts at labor reform, but it is also seen by stakeholders as deeply troubling in its lack of transparency and due process. Through its immediate penalty system, the RRM empowers the initiating government to shoot first, figuratively, and ask questions later. As we explain further below, the Mechanism is also asymmetric such that, practically, only the United States and Canada can use it toward Mexican firms.

Our aim in this Current Development is to analyze the first uses of the RRM and to provide a preliminary assessment of its contours. Through September 15, 2023, the U.S. government had invoked the tool on thirteen occasions. We offer a comprehensive empirical review of the situations that are publicly known, and some that are not available in the public domain. Our study shows that the use of the RRM gives rise to questions about its consistency with other principles of international law. We consider the political economy of the RRM, and argue that the Mechanism, while a noteworthy advance in transnational corporate governance, has mixed repercussions for its constituencies, including those it purports to help the most. It has proven useful to certain workers in Mexico as they have sought representation in their companies, but it suffers in operation from the selective and non-transparent methodologies used by the governments.

As some U.S. lawmakers now refer to the RRM as the baseline for corporate accountability tools within economic agreements,¹² and with an eye to the new RRM-styled tool recently announced as part of the Indo-Pacific Economic Framework for Prosperity (IPEF),¹³ we review lessons learned so far for such a model and consider its future.

I. The Mechanism in Context

The creation of the RRM turned on an extraordinary political moment in the United States and among its neighbors. A confluence of a rise in populist ideology and the evolution of the North American auto industry facilitated this new international institution. This Part unpacks that story before turning to a review of how the RRM works.

A. Getting to Yes

Understanding this rare moment of convergence, which was the product of exceptional political, legal, and socioeconomic features in the two prior years, requires some background on one of the focal points in the renegotiation of the North American Free Trade Agreement (NAFTA): highly integrated supply chains in North America. Among the most important of those supply chains is that concerning automotive goods. The automobile sector played a disproportionately large role in the NAFTA renegotiations and, as will become clear below, a significant role in the initial rollout of the RRM.¹⁴

Before the Trump administration came into office, its leading economic policymakers were preparing for a renegotiation of the NAFTA to address grievances in the automotive sector.¹⁵ Candidate Donald Trump had run in 2016 on a platform opposed to trade with Mexico and the NAFTA, including the threat of 35 percent tariffs on products from Mexico.¹⁶ By fall 2017, the NAFTA renegotiations were underway and were heavily influenced by then-President Trump's rhetoric, including his threats to “rip up” the NAFTA¹⁷ and to impose tariffs on cars,¹⁸ as well his stance on immigration.¹⁹ A variety of contentious issues in U.S.-Mexico relations came to the fore with considerable discontentment on the minds of U.S. lawmakers. This created tremendous uncertainty, especially for Mexico, which the United States used as leverage, given Mexico's massive dependence on the U.S. market for its exports.

The renewed scrutiny on the North American region and its economic nuances coincided with the enforcement of—and dissatisfaction around—international labor law under the NAFTA and other U.S. economic agreements. Longstanding dissatisfaction arose from a perception that trade agreements had failed to achieve measurable improvements for workers.²⁰ Since the NAFTA, the United States had included binding and enforceable labor rights provisions in its trade agreements either as side agreements or embedded in the agreement.²¹ Those provisions committed the parties to maintain certain labor standards in their domestic law, consistent with some of the International Labor Organization conventions, and to effectively enforce those standards.²² They often included means through which the public could raise concerns with the governments about a trading partner's labor enforcement practices. Although the U.S. government had facilitated some labor reform in partner countries through these trade agreement provisions, the absence of tangible enforcement efforts raised questions for progressive proponents.²³ Despite receiving several public communications concerning violations of labor rights in trading partner countries through these mechanisms, as of 2017, the United States had only undertaken one state-to-state labor enforcement action.²⁴ That arbitration against Guatemala began nearly a decade earlier, under the Dominican Republic-Central America-United States Free Trade Agreement. The arbitral panel concluded that Guatemala had not breached the labor chapter of the trade agreement despite having found that Guatemala had failed to effectively enforce its labor laws.²⁵

With the news of the Guatemala decision breaking just as negotiations began on an updated NAFTA, support for the once-heralded integration of labor provisions into economic agreements quickly faded. Addressing Mexico's broad labor exploitation problem became pivotal to the NAFTA renegotiations. For many years, the United States and other governments have put pressure on Mexico to address the prevalence of exploitative and undemocratic unions, known as “protection unions” [“charro” unions]. When the three governments signed a North American economic agreement in November 2018, it included requirements for Mexico to revise its worker protection laws and institutions to end protection union dominance. But attention to Mexico's labor issues did not end with the conclusion of the new agreement.

The November 2018 election gave Democrats a majority in the U.S. House of Representatives. The new North American agreement had not yet received congressional approval and the newly empowered Democrats in the House of Representatives demanded changes, including stronger labor commitments. Negotiators experimented with different models that would go beyond merely requiring trading partner commitments and relying on ordinary state enforcement channels. Earlier drafts floated a company certification system and an approach modeled on the U.S. Foreign Corrupt Practices Act.²⁶ The U.S. delegation offered proposals based on prior tools like that in the U.S.-Peru Trade Promotion Agreement of 2009, designed to combat illegal deforestation,²⁷ and another from a 1999 agreement with Cambodia policing labor violations in the textiles industry.²⁸

In December 2019, the parties reached a deal (hereinafter “the Agreement”).²⁹ The final result of these labor-driven additional negotiations was a new annex (the RRM) to the dispute settlement chapter that shifted liability for worker rights’ violations away from the state and instead put penalties on companies. Put very simply, the tool enables one of the parties to enforce penalties, typically through trade restrictions, on a worksite on the territory of another party where workers are denied the right to bargain collectively. This major development appeased much of the strong, organized labor community in the United States.³⁰ Reflecting on these additions, one former negotiator referred to the 2019 updated Agreement as more of a labor agreement than a trade agreement.³¹ The United States’ success in securing this arrangement is a tremendous development for CSR and its related movements: this new North American economic integration treaty codifies transnational enforcement of worker rights.

B. Mechanics

As a matter of law, the RRM offers several paths to reach the same end: ensuring workers are able to freely unionize at their places of employment. But significant aspects of the tool remain open to broad interpretation by the parties and stakeholders. The Mechanism is structured around four junctures: (1) the initiation of a review of a worksite; (2) the implementation of a course of remediation where the governments identify an ongoing denial of rights; (3) the possible establishment of a review panel (upon request); and (4) the imposition of remedies. We leave a comprehensive overview of the Mechanism's many facets to our technical writing where we have walked through them with greater precision.³² Here, we present the basic configuration of the tool and those elements that make the RRM unique and transformative. What is most important for our international legal analysis is the institutional framework for this firm-centric tool and its transnational features.

The RRM can be initiated by any of the three governments when that government has a good faith belief of a “denial of [collective bargaining or freedom of association] rights” at a worksite on the territory of another party.³³ That “belief” typically relies on information from members of the public. Consequently, the Mechanism features a significant public-facing component; all three parties have provided means through which members of the public may submit information about labor rights situations, primarily through petition or through a web form “hotline” where information can be reported anonymously.³⁴ One of the governments can activate the RRM regarding a facility in a “priority sector,” defined as “a sector that produces manufactured goods, supplies services, or involves mining”³⁵ if the facility produces goods or supplies services traded between the parties or that compete in the territory of the other party.³⁶

At the opening stage, the initiating government is entitled immediately to “delay final settlement of customs accounts related to entries of goods” coming from that facility.³⁷ If the territorial government is uncooperative, or where the parties disagree, either government may call for the establishment of a neutral panel of labor experts to determine whether workers are being denied their rights.³⁸

Where the parties jointly believe a denial of rights has taken place or is ongoing, they may agree to pursue a “course of remediation.”³⁹ The Agreement does not offer specific content for the course but those to date have included, for example, requirements to hold a new union vote and for that election to be monitored by external observers. Where no course of remediation is concluded or successful, the initiating government can decide to impose penalties on the company until the denial is ameliorated.⁴⁰

C. International Law Innovations

Even before its first use, the RRM on paper has five features that make it especially noteworthy among international law instruments: its open-ended penalty system; its hybrid labor/human rights logic; its mix of compliance and dispute settlement features; its cross-border reach; and, its asymmetry.

First, the RRM text offers a non-exhaustive list of possible penalties for companies that do not successfully remediate the denial of rights at their facilities. The Agreement is structured around the suspension of final customs processing of inbound goods at the border; that preliminary step occurs early in the review process. Thereafter, the company risks loss of preferential tariff treatment for the goods made at that worksite,⁴¹ and in the case of repeat offenders, the initiating government may deny entry to all the goods from the company,⁴² limited only by imprecise principles of proportionality.⁴³ It is less clear whether the governments have leverage to take such forceful action on a facility that does not export its wares directly. For example, no indication is given as to whether or how the governments would consider stripping a product (e.g., tires) of its tariff-free status when it is integrated into another good (e.g., an assembled vehicle), or just how far down the supply chain the ramifications may fall. Likewise, for facilities providing services rather than goods, the Agreement uses only the word “penalties” without any indication of what penalties would be consistent with the Agreement and with other tenets of domestic and international law. In short, the text gives the parties considerable discretion in taking action toward companies.⁴⁴

Second, the RRM focuses on “remediation” of the denial of collective bargaining rights, principally by repeating or reorganizing collective bargaining elections. In other human rights contexts, governments have developed a more robust meaning of remediation: one that focuses on making workers whole. For example, the EU is presently considering a proposed regulation to compensate and otherwise provide retrospective relief to those harmed by forced labor.⁴⁵ In that context, companies may be required not only to cease forced labor practices but also to provide financial and other support to victims.⁴⁶ In some RRM situations, the governments have required that companies reinstate and compensate wrongly dismissed workers.⁴⁷ Apart from individual instances providing such relief, however, the RRM does not on its face suggest it is in the business of compensating rights victims.⁴⁸ The tool is built to offer remedies that are prospective and not punitive, and that cease when the issue is resolved. Trade remedies laws such as anti-dumping and countervailing duty laws work in a similar way, as does authorized retaliation in a dispute under the World Trade Organization. Using this approach in a worker rights context lays bare a tension between the instrument's human rights aspirations, seeking to mediate between individuals and the state, although notably without human rights protections itself, and its labor rights reality, intervening between labor and capital. While the tool implicitly suggests some failure on the part of the territorial state, by imposing penalties on the company rather than the government, the RRM shifts responsibility for remediating the harms rendered on the workers to the company.

Third, the RRM does not fit easily into pre-existing categories of international law enforcement tools. If anything, it is more appropriately considered a compliance mechanism than a dispute settlement mechanism, and this distinction has consequences beyond just precision and rhetoric.⁴⁹ Nevertheless, two features of the Agreement have led commentators to treat the Mechanism like a dispute settlement tool between the two states. First, the Mechanism appears in a state-to-state agreement as an annex to the dispute settlement chapter and, second, the text refers to the governments as the “complaining Party” and the “responding Party,” just as one would expect in a dispute settlement context. But the state parties may use the RRM without the presence of any dispute. In several of the RRM situations, the state parties have agreed about the denial of rights. Where a panel is established, however, such as in an August 2023 situation involving Grupo México, the panel is called upon to make a factual determination, rather than reach a legal conclusion about party consistency with the Agreement. As a result of these mixed terminologies and experimental processes, the RRM adjusts the levers of control in the conventional compliance story. The incentive structure of the RRM does less to constrain unlawful state behavior; rather, it serves as a greater constraint on private behavior. Moreover, it does so without delegating that authority to an intergovernmental organization, and the panels that it foresees do not operate like traditional international dispute panels.⁵⁰ Thus, unlike first-order state responsibility for breaches of inter-state agreements, and wholly apart from second-order compliance in the form of consistency with adjudicatory decisions, this tool creates a third-order compliance regime that seeks to replicate the domestic enforcement chain through a transnational instrument.⁵¹

Fourth, the reach of the RRM exceeds most conventional boundaries of transnational regulatory monitoring and enforcement. The RRM text permits parties to undertake work that the labor chapter of the same Agreement expressly prohibits: they may enforce domestic law on the territory of another.⁵² It creates a diagonal enforcement exercise according to which the initiating government acts directly against the company in the other territory. Contrast this move with the horizontal enforcement between state parties in ordinary trade agreements; most U.S., Canadian, and Mexican trade agreements empower their state parties to enforce the agreement against one another. And contrast it with vertical enforcement carried out by domestic regulators of their own law within their own jurisdictions, or even abroad. The RRM provides enforcement at the border, but it otherwise parallels few other trade measures. The RRM has similar reach as forced labor bans, for example, and food safety rules by enabling states to investigate individual worksites and evaluate them according to standards developed by the U.S. government concerning Mexican labor law. Food safety rules also empower governments to stop plants or foods at the border that, further to scientific review, may pose a health risk to the population; denials of collective bargaining rights can hardly match that degree of public harm but the companies engaged in those practices face a similar outcome. Further, in RRM situations, one government is interpreting and applying another's labor law, not the text of the Agreement or its own law. The foreign officials go to the worksite to investigate each situation. Making matters more complex, although the text of the RRM creates no new obligations for private actors, some of the remediation plans Mexico and the United States have developed in these first three years have imposed new obligations on the implicated company. Those new obligations may be imposed without notice, and the Mechanism provides no opportunity for administrative or judicial review.

Fifth, the tool, which is agreed between Mexico and the United States on the one hand, and between Mexico and Canada on the other, is reciprocal in that the primary text applies to all three governments, but in practice, limitations added to its application by the United States and by Canada create significant asymmetry. There is effectively no opportunity for Mexico to initiate an RRM about a U.S. facility because a claim can only be brought with respect to a facility “under an enforced order of the National Labor Relations Board [(NLRB)]”⁵³—a category invented for the Agreement. This carve-out has raised questions about the legitimacy and legality of the tool.⁵⁴

II. Situations to Date and Their Impacts

Through September 15, 2023, the United States had invoked the RRM thirteen times to intervene at a worksite in Mexico. Canada accepted one petition for review and resolved the issue before invoking the RRM.⁵⁵ Mexico had not initiated any RRM situations; as noted above, the asymmetry in the tool makes it unlikely that Mexico will ever use it.

Below, we provide a 360-degree review of the RRM situations to date, explaining the labor violations, actions taken by each government, and results of their engagement. We also explain the actors behind the petitions that have been filed, what their interests are, their choices among worksite targets, the companies that are affected, and the place of those companies in the automotive supply chain.

A. Overview: Dates, Numbers, and Worksites

Somewhat surprisingly, none of the three governments invoked the RRM until nearly a year after its entry into force. The United States did not use the tool at all during the Trump administration and did not engage on the RRM until several months into the Biden administration.⁵⁶ In the subsequent two (plus) years, however, the U.S. Trade Representative (USTR) has, on thirteen occasions, sought review by Mexico of an alleged denial of rights at a worksite in Mexico. Table 1 names the companies, the locations of their worksites, and the dates on which the United States sought Mexico's review. Mexico accepted ten of those requests for review and ultimately rejected three—two because the alleged denials occurred before entry into force of the agreement and one because there was no evidence of employer interference or denial of rights to freedom of association and collective bargaining by the company.⁵⁷ The table also provides basic data on the rejected petitions about which some limited information was available. Almost no information is publicly available about the petitions that are filed and rejected.

Table 1. Information on USMCA RRM petitions and situations, July 1, 2020–September 15, 2023

Each situation, as well as the rejected petitions publicly known, dealt with a facility in the automotive sector until June 2023 when the United States sought review of a facility in the textiles sector. Thereafter, the United States also sought review of facilities in the mining and services sectors. Observers have acknowledged that, as explained above, the auto sector is undoubtedly one of the most important for the region, along with agriculture, which is not included as a priority sector under the RRM.⁵⁸

All but two of the thirteen situations were triggered by the filing of a petition by a non-governmental organization or labor union. One was commenced following a tip to the U.S. hotline. In its earliest days, U.S. officials emphasized that they planned to “self-initiate” many RRM situations, using the vast resources provided to them under the USMCA Implementation Act.⁵⁹ Despite that promotional language, the U.S. government has initiated only a small handful of RRM reviews based on its own activities and engagement in Mexico.

B. First Tests

The first invocation of the RRM by USTR involved a subsidiary of the major U.S. auto manufacturer General Motors (GM).⁶⁰ By 2018, GM had become Mexico's largest producer and exporter of automotive vehicles, operating three different final vehicle assembly plants in addition other facilities located across the country.⁶¹ On May 12, 2021, USTR requested that Mexico review whether a denial of rights was occurring at the General Motors de México facility in Silao, State of Guanajuato, about four hours northwest of Mexico City. GM's Silao facility, which spans 602 acres, and includes facilities for propulsion systems assembly, vehicle assembly, and stamping and paint shop, began operations in 1995.⁶² The popular GMC Sierra and Chevy Silverado trucks are among the vehicles manufactured in Silao and exported to the United States by more than 6,000 workers.⁶³

USTR's review request, which was premised on information it received through the confidential hotline, raised concern that workers at the facility were being denied the right of free association and collective bargaining during an April 2021 vote for approval (“legitimación” or “legitimization”) of a collective bargaining agreement between General Motors de México and the “Miguel Trujillo López” union. Upon issuance of USTR's request for review, Trade Representative Katherine Tai directed the U.S. Treasury to suspend liquidation of “all unliquidated entries of goods” from the Silao facility, which had the effect of pausing the final processing of the charges associated with importing the trucks into the United States. (This could have been economically consequential, given the 25 percentage-point differential between the tariff that could be imposed and the preferential tariff rate for trucks under the Agreement.)

On July 8, 2021, the United States announced together with Mexico a “course of remediation” that would remediate, in their view, the denial of the right of free association and collective bargaining.⁶⁴ It included requirements such as holding a new vote and monitoring by Mexican inspection personnel and observers from the International Labor Organization.⁶⁵ The remediation plan also required GM to take steps to investigate any instances of retaliation or reprisals against workers. A new vote was held in August 2021. According to press reports, workers at the plant rejected a previous collective bargaining agreement and ousted the dominant union.⁶⁶ A total of 5,876 GM employees cast ballots, with 3,214 workers rejecting the bargaining agreement and 2,623 workers electing to keep it.⁶⁷ On September 21, 2021, U.S. authorities announced the successful conclusion of the remediation plan and resumed liquidation of the Sierra and Silverado trucks. Later, a new union was formed and bargained collectively to negotiate a wage increase of 8.5 percent for workers at the plant.⁶⁸

Already in the first deployment of the tool, the GM situation illustrated some of the challenges such a tool can and has posed. In the GM situation, both the U.S. government and the Mexican government were heavily involved, including outside the boundaries of the RRM. Prior to the U.S. invocation of the RRM, the voting process had been suspended by Mexican government officials due to irregularities including the destruction of ballots, and potential violations of Mexican laws. Mexican authorities had filed a criminal complaint about individuals involved in the ballot tampering.⁶⁹ And, the day before USTR's review request to Mexico, the Mexican labor authorities ordered the dominant union at the plant to hold a new vote within thirty calendar days.

In this instance, the RRM served as an additional compliance regime on top of the work underway by the Mexican labor ministry and other Mexican government officials. But the RRM also changed the landscape of the responsible and penalized parties. Prior to the U.S. engagement, Mexican officials had taken action against the exploitative and undemocratic union, as well as against employees implicated in the ballot tampering. The effect of layering the RRM over this domestic enforcement was to add immediate financial constraints and possibly additional liability as more facts came to light from the review. Put differently, the RRM served to enlist the company in also facilitating the desired outcome of remediation of worker rights to democratically elect a union.

Subsequent RRM situations followed a similar pattern. With respect to a Panasonic facility, for example, information released by USTR stated that the company entered into an illegitimate collective bargaining agreement with a protection union, leading to dismissal of workers and withholding of paychecks. The facility ultimately agreed to renounce a collective bargaining agreement it had with a protection union, reimburse workers, and offer backpay to workers who were wrongfully dismissed, among other commitments.⁷⁰ At a VU Manufacturing worksite, the United States alleged that the company cooperated with an illegitimate union without workers’ consent.⁷¹ This invocation of the RRM led to elections at the worksite carried out with government supervision.⁷² Four months later, however, members of the U.S. Congress urged the USTR to take further action at VU Manufacturing, alleging that the illegitimate union had made threats toward a labor organizer and that violence at the facility was likely imminent. This led to a second activation of the RRM against the same facility, resulting in a course of remediation that is scheduled to be completed by the end of September 2023.⁷³

In August 2023, the USTR announced that it sought the establishment of a panel for the first time under the RRM for a situation involving a Grupo México regarding a mining facility, after the Mexican government rejected the U.S. request.⁷⁴

C. Parent Companies

All but two of the targeted facilities are Mexican subsidiaries of foreign-headquartered, multinational firms. The use of the tool is based on the location of the facility, not company nationality.⁷⁵ Seven of the situations initiated by the United States have involved U.S.-headquartered firms, two have implicated Japanese firms, two have targeted European companies, and the situation reviewed by Canada involved a facility with a European parent company. Among the known rejected petitions, at least two have involved facilities that are U.S.-owned and one has a parent company in Europe. The RRM's facility-specific nature extends its reach to regulate companies far beyond the borders of Mexico, the United States, and Canada. Other governments are taking note, but not yet publicly intervening on behalf of their multinational companies.

Some commentators expected that the Mechanism would be used mostly on large auto manufacturers at their Mexican facilities, and that that could have created a broad deterrent effect on employers throughout the auto supply chain. Yet, nearly all the situations since GM have occurred at much smaller automotive part facilities, most with little brand recognition. Moreover, the automotive parts that they manufacture often have close substitutes. Whereas no other company can make a Sierra truck, many can make arm rests to go in those trucks. In those circumstances, intervention by the United States could jeopardize the reliance of the big companies on those smaller suppliers and in turn put those workers’ jobs at risk. This supply chain relationship may limit the power of the workers to seek to use the tool even where they experience a denial of rights. On the other hand, attorneys in private practice report that some companies have taken note and changed their behavior to avoid RRM attention, and labor movement advocates note that workers have felt emboldened by the RRM to advocate for their rights even without filing a petition with the U.S. government.⁷⁶

The RRM also poses a challenging legal question about attribution: the penalty is directed toward the company when the company may not be at fault. In at least some of the RRM instances, the actors forcing the denial of rights have been other employees as opposed to company management. Commonly in Mexico, the denial of rights at least in part is generated by the existing protection union, sometimes with the support of local government officials.⁷⁷ The RRM in some respects expedites Mexico's national reform effort by putting pressure on the company, rather than the state or civil society, to address any rights violations regardless of the source.

For facilities that export to the United States, the RRM imposes penalties immediately before the completion of any investigation into the alleged denial of rights. Despite this strict regime, the role of the company in the RRM is obscured by its absence in the text of the Mechanism and also in the practice. Nothing in the RRM requires the three governments to involve the company in its review or in its subsequent remediation planning and rarely have they done so. Participants in the RRM process have indicated that the parent companies are often not aware of the alleged denial of rights prior to the commencement of the RRM review.⁷⁸ In the remediation plans that are publicly available, the two governments have repeatedly included several obligations on each company to ensure free and fair union elections.⁷⁹ Company representatives have reported that the governments have not consulted them in the development of these plans,⁸⁰ and other stakeholders have likewise raised concern about their exclusion.⁸¹ Matters are made still more complicated by the distanced role of the company in the collective bargaining process. Companies are prohibited from interfering with the union-forming process and may consequently operate with limited information about its operationalization.⁸² Where the RRM demands company engagement, companies now need to navigate that fine line.

D. Unions and Geography

As suggested above, the problem that the RRM most often tries to solve is to allow workers to vote for union representation whereas, throughout Mexico for many decades, the labor movement has been dominated by exploitative unions that have not been freely elected. Nearly all the efforts undertaken under the RRM so far seek to displace unions associated with the national Confederation of Mexican Workers (CTM), long considered to be dominated by protection unions.⁸³ Thus, the primary use of the RRM has been to remove the unfair or abusive union rather than enable workers to unionize in the first instance.⁸⁴

Table 1 captures the heterogeneity among the unions and civil society actors that have brought RRM petitions. Although no one actor dominates the filings, the United Auto Workers and the United Steelworkers, two of the most powerful labor groups in the United States, as well as the American Federation of Labor and Congress of Industrial Organizations, have been instrumental in the development of petitions for review, as have non-governmental organizations such as Rethink Trade. Likewise, on the ground in Mexico, labor leader Susana Prieto has advocated for the use of the RRM and her activism has likely contributed to the concentration of RRM activity in the auto sector and at worksites in particular regions.⁸⁵

Figure 1 also offers a map of worksites targeted. It shows that some cities have been subject to more than one review, in part because workers may have shared information about the RRM process with their neighbors and local communities.⁸⁶ These locations are some of the major automotive manufacturing centers in Mexico.⁸⁷ And, several of the Mexican states that are home to the facilities are states that have been slower to adopt the labor reforms.⁸⁸

Figure 1. Mexican facilities subject to Rapid Response Labor Mechanism petitions or situations between July 1, 2020 and September 15, 2023.

Source: constructed by the authors, map data © OpenStreetMap. With the exception of INISA 2000 (textiles), Grupo México (mining), and Mas Air (cargo transportation services), all facilities are in the automotive sector.

E. Speed

A key component of the RRM is that it is “rapid.” Compared to the U.S.-Guatemala case, nearly any proceeding would be faster, but just how rapid is the RRM? Precise start and end dates are difficult to measure. If we measure from the date of the original petition to the date that liquidation is resumed, each of the situations encountered to date has lasted fewer than four months. They range from 81 days in the case of Unique Fabricating to 132 days for GM.

These numbers undercount in one respect and overcount in another. First, the date of resumption of liquidation is not the end of the matter. In each situation, monitoring by both governments continues for at least several more months through the implementation of the remediation plan⁸⁹ or through the informal, unannounced supervision that both Mexico and the United States have undertaken.⁹⁰ In each instance, USTR has provided subsequent status reports, celebrating wage increases achieved by new unions after the investigation concludes.⁹¹ Second, the date of the petition may not reflect the date upon which the U.S. government's review began, and with respect to both GM and Draxton, no information is available regarding the date of the hotline tip or other investigation.

Finally, that VU was the subject of a second petition complicates any analysis of completion or success. Was a second petition required, and what would its legal consequences be, given the Mechanism's enhanced penalties for repeat offenders?

III. Assessments and Challenges

In prior decades, the introduction of labor commitments into economic agreements reflected a cacophony of rationales and basic principles.⁹² Different theories buttressed the idea that trade and labor should be linked: protectionism, work protection, fairness, development, human rights, the absence of any other labor enforcement mechanism, and others.⁹³ While scholars continue to speculate about the trade-labor narrative, the governments of North America have stepped away from those debates to focus greater attention on corporate actors and their specific actions, rather than on broad policy statements about stringency and reform.⁹⁴

Viewing the RRM as a corporate accountability tool rather than a trade or trade-and-labor enforcement mechanism yields new theoretical and functional considerations. At the most practical level, the RRM creates costs for several constituencies as much as it may lead to benefits for those same groups. This Part examines those positions relying on interviews with key players in the development and implementation of the RRM.

Outwardly, the United States hails the RRM as a success in achieving or unlocking higher wages and other benefits for workers not only at the eight sites where the United States has taken action but also at other sites where a petition was imminent or where U.S. personnel collaborated closely with a union and other relevant actors.⁹⁵ Internally, however, some bureaucrats harbor a degree of uneasiness about the overall value of the RRM. The RRM process is time consuming for many staff, requires extensive government resources, and may not lead to any measurable net benefits.⁹⁶ The relatively small number of situations initiated to date is due in part to these administrative burdens.⁹⁷ Members of the U.S. Congress have sought increased use of the tool and demanded more action from USTR.⁹⁸

Mexican policymakers take varied positions. At present, the left-wing López Obrador administration supports strong labor enforcement and the broader labor reform required by the Agreement.⁹⁹ Undertaking that national level reform is a huge task that cannot be completed in the short time afforded under the Agreement. One view of the RRM is that it helps the Mexican government achieve greater reach in its enforcement efforts. Additionally, at the sub-federal level, not all Mexican states are cooperative partners in the reform effort. The RRM may have proven useful in intractable regions. Finally, from Mexico's perspective, the shift in design to now focus on corporations rather than to target the government is a beneficial reconstruction and redistribution of costs on actors that can more easily bear them.

Despite these gains, not all deployments of the tool have occurred with the López Obrador administration's consent. In the case of Tridonex, the facility targeted in the first petition filed with the U.S. government, the Mexican government declined to accept the petition for review on the grounds that the events detailed therein pre-dated the entry into force of the Agreement, putting them outside the scope of the RRM. The Agreement provides that, where the receiving government declines to accept the review, the initiating government can convene an expert panel if it maintains its belief about a denial of rights. But following Mexico's refusal, USTR contacted the U.S. parent company of Tridonex directly to resolve the matter.¹⁰⁰ USTR devised a bespoke remedy, outside the four corners of the Agreement text, and entered into an arrangement with the company to remediate the denial of rights at the facility.¹⁰¹

The U.S. and Mexican governments have mostly cooperated on the RRM situations thus far. As noted, the Mechanism has support from left-leaning administrations in both the United States and Mexico. When those administrations change, however, the governments’ deployment of the RRM could play out very differently, and perhaps more aggressively. In an over-sanctioning scenario, the RRM could become unsustainable and its bite scaled back as in other international economic law dispute settlement contexts. Further, if future administrations are not sensitive in such a way that facilitates North American economic integration and serves Mexico's own interests, policymakers will find it difficult to support its maintenance. Expanded interference could be perceived as infringing on party sovereignty that could in turn put the North American economic integration project in doubt.

The non-governmental organizations and organized labor groups that have submitted petitions for review have indicated, positively, that the RRM is supporting stronger corporate accountability norms and better conditions for workers in the locations where it has been used. But some have lamented that the tool has not led to sufficient institutional capacity-building in Mexico, putting significant onus on U.S. organizations.¹⁰² Just as the U.S. government officials have found the review process laborious, the filing groups must engage in extensive research and writing to submit a single petition regarding just one facility. They have done so with little guidance from U.S. officials as to the criteria for review. The lack of transparency by the governments has created a strain on organizational resources.¹⁰³ Further, some advocates have noted that achieving higher wages for workers appears at first glance consistent with labor advocacy goals but they likewise acknowledge that they lack sufficient information to evaluate the gains of the RRM. If misused, the tool could make matters worse by negatively impacting the workers it is designed to help, or by disrupting local communities and contributing to power imbalances that may lead to increased violence in those communities.

Finally, the business community harbors mixed views about the tool. For some targeted companies, learning about denials of rights, including those that violate domestic law, is critical to shoring up their supply chains. The initiation of investigations by the U.S. government has prompted companies to take a closer look at the management of their local subsidiaries. Targeted companies have carried out their own investigations to shield themselves from further liability and to address informational disparities arising in their conversations with the U.S. executive branch, the U.S. Congress, and other sources. Those disparities have also yielded strong concerns about the lack of transparency in the RRM and the companies’ exclusion from the remediation conversations. They are not “full partners,” although one U.S. official has suggested that full partnership was the intention behind the Agreement.¹⁰⁴ They have sought regular meetings with USTR but those have yielded little information. As one business representative put it, “they wanted info from us and we wanted info from them,” without any clear understanding about what the other was seeking.¹⁰⁵

Some of the smaller companies under review could face significant financial distress, depending on how long it takes for the United States to order liquidation to resume. Companies risk a cash flow problem during the suspension of liquidation that may require the posting of a considerable bond for an indefinite period. Although neither the Mechanism nor the statute adopting the Agreement in U.S. law provides clear recourse to judicial review, companies may elect to seek relief in domestic courts or through investor-state dispute settlement.

IV. Corporate Accountability's International Law Moment

The early operation of the RRM as the first binding and enforceable international corporate responsibility mechanism yields lessons, particularly with respect to drafting legal language and for different design options. The time is ripe: in August 2023, fourteen countries shared draft text of the IPEF Supply Chain Agreement which includes a facility-specific labor mechanism.¹⁰⁶ Elsewhere, governments have discussed extending the RRM to more labor rights and to police carbon and climate change obligations imposed on companies.¹⁰⁷ The European Union is finalizing its corporate due diligence directive.¹⁰⁸ Considering these lessons is now a matter of urgency.

Our study demonstrates that at the top of the reform list is addressing the critiques about the rule-of-law deficiencies and lack of due process for affected companies and for petitioners, and not only for reasons of principle. Addressing worker rights requires the full cooperation of all affected parties. If policymakers seek to adopt an RRM in future contexts, they ought to restructure company engagement, ensure greater transparency, and provide extensive public guidance on how to use the tool. None of the stakeholders, including the governments, has known how to demonstrate credible evidence of a denial of rights in RRM processes. After more than two years of review, the United States in 2023 issued limited guidance on its review of petitions which largely did not resolve stakeholders’ outstanding questions.¹⁰⁹

Another area for fruitful reform is more elaborate text explaining what the parties intend with respect to the verification opportunities for panels and an elaborated list of remedies or paths to remediation. With the current text, panels have little guidance on their work and the scope of their investigation—an issue already highlighted by the first panel now underway—and companies have even less on which to rely to preemptively resolve matters that would help them avoid liability.

A review of the operation of the RRM to date could generate alternative design options. Rather than alter the landscape with the unions, the governments could educate the companies and the workers as to the law and create other pathways to rapid resolution. Governments could consider a “trusted trader” program or “safe harbor” program for companies that have met labor standards goals.¹¹⁰ They might institute an auditing program with a structure that resembles the Committee on Foreign Investment in the United States where a joint commission could review situations with a credible basis to believe there has been a denial of rights and then limit the company's ability to export goods until the company provides a satisfactory report about the situation.¹¹¹

The RRM also offers ideas for new theoretical constructs of transnational compliance. That states will agree to police each other's domestic laws vis-à-vis private actors, including among those entities with no jurisdictional link to the territory of the enforcer state, opens new inquiries into basic foundations of state responsibility, compliance, and regulatory cooperation.

The RRM also contributes to the development of labor law. Although it does not go so far as to create an international common law of labor, the parties make determinations about whether facilities are operating consistently with Mexican law, creating a type of internal precedent. These actions affect firm behavior. The remediation plans have specific requirements for how to manage union votes through which they effectively write hard and soft law. These requirements, formulated by the United States and Mexican governments, fill gaps in Mexican law. Thus, even though the RRM does not create new generally applicable obligations, it reinforces domestic labor law and may alter its application.

The RRM is both broad and narrow. Its power turns on cross-border supply chains, though its legal reach is not so limited. Governments can target nearly any facility in the export or internationally competitive sector. And yet, the RRM is focused exclusively on Mexican workers’ inability to bargain collectively, a problem stemming from a history unique to Mexican labor law and experience. The RRM does nothing on its face about other rights’ violations, such as workplace conditions, discrimination, or ensuring workers are paid a minimum wage. At best, by empowering workers and allowing them to unionize—and bargain collectively—this may allow them to advocate for additional rights they want to prioritize for themselves.

The RRM got off to a slow start. Its use is now accelerating, with eight initiations in the first nine months of 2023 alone. Two critical benchmarks are likely to affect its future: first, the broader Mexican labor law reform. The deadline for legitimizing new union contracts has passed with few contracts in compliance.¹¹² The RRM now faces its most significant test: will it pick up that slack? Second, the Agreement has a sunset clause, and that sunset is not far away. In 2026, the parties will review its operation, including that of the RRM. Between now and then, RRM-fatigue may set in and the politics in North America are surely to change again.

Conclusion

Although the RRM's purview is limited and the data so far are few, tracing the RRM experience benefits broader discussions about corporate accountability in international law. The Mechanism extends beyond traditional soft law corporate accountability tools, notably by reaching beyond corporations to other exploitative actors and to gaps in state enforcement. As CSR scholars have acknowledged, challenges arise any time states seek to negotiate agreements that insert states into relationships among other stakeholders—here, unions, businesses, and communities.¹¹³ That is even more so given the RRM's binding and blunt nature.¹¹⁴ While other corporate accountability tools operate through back door, the RRM enters through the front door. And it does so by prioritizing labor rights advances over procedural norms. This experimental instrument is shaking up prior understandings, and what we have seen so far in the North American context may be just the beginning.