Policy recommendations

  • With sustained and vigorous efforts from trade unions, labour provisions in trade agreements have been of some help to labour movements in some partner countries, but they could do much more if the lessons from recent examples are truly learned.
  • Labour provisions should relate to the immediate needs of workers in a partner country and go beyond the fundamental conventions of the International Labour Organisation (ILO).
  • There should be provision for rapid sanctions, as these do work. While moral condemnation can strengthen trade unions in a partner country, provision for commercial sanctions is proving much more effective.


Labour provisions were included in 85 of the 293 free trade agreements (FTAs) in force in 2019, but a common verdict has been that they made little or no difference to labour rights and working conditions (Myant 2017). This is despite the hopes of the European Union (EU) and United States (US), the driving forces behind their development, that they would help spread good practice and prevent unfair competition from countries that do not respect labour rights. In fact, the record has been unimpressive. Disputes have been very few and far between: by 2019, only one had been initiated by the EU (ILO 2019b: 55). The US was more active, but enforcement by imposing fines or ultimately trade sanctions, an eventuality provided for in US agreements, seemed impossible after the notorious US-Guatemala case, concluded in 2017.

This case was the only one at the time to run the full complex course, taking nine years from the trade unions’ submission of a complaint to the final verdict by the required panel of experts. The undeniable violations of labour rights in Guatemala did not satisfy the stringent conditions set in the FTA (of being both ‘sustained and recurring’ and giving a trading advantage) required to justify imposition of sanctions (Ortino 2021). The EU’s method of lengthy dialogue and a dispute procedure that ends without sanctions also appeared ineffective.

The limitations in labour provisions stem ultimately from the reality that trade agreements were driven primarily by business interests, with labour provisions included as a concession to ensure backing across the political spectrum. That means some account was taken of the voice of labour by the EU and the US, but employment issues were not pressed as priorities during the negotiation of agreements and nor were they given the same weight as commercial issues after agreements had been signed (Smith et al. 2020). However, some recent cases show that labour’s voice can make some difference, if it is persistent and supported internationally and if conditions are favourable in partner countries. The recent developments in relation to South Korea, Georgia, Vietnam and Mexico are covered in detail in an accompanying ETUI working paper (Myant 2022). They are summarised below.

South Korea

The EU-Korea FTA came into effect on 1 July 2011. The Korean government was keen to promote trade agreements and, as a developed country with per capita GDP close to the EU average level, it was in a strong enough bargaining position not to need to accept all the EU’s proposals. For example, it did not accept a requirement to ratify ILO fundamental conventions on rights to free association and collective bargaining which were opposed by employers. It accepted only a commitment in a sustainable development chapter to ‘respecting, promoting and realising’ the rights they embodied.

Korean employers’ favoured model of industrial relations is based on loyal company unions. However, this began to be challenged in the 1980s by the rise of more militant unions, which formed, in 1995, the Korean Confederation of Trade Unions (KCTU). This confederation won legal recognition only after the government’s repressive measures had been met with strike action. Even after that, Korean trade unions experienced a variety of forms of victimisation and harassment, periodic imprisonment of leaders, restrictions on the right to strike, delays over and barriers to registration of unions, and restrictions on who could hold union office. These issues were raised repeatedly by the KCTU and international trade union organisations to the ILO, which could only conclude, on the basis of the evidence before it, that Korean employment law was not compatible with its fundamental conventions.

The main force for change in Korea has always been internal politics, and a government somewhat more sympathetic to trade union demands came to power in 2017. The EU was a help to those pressing for change after being persuaded to recognise the views of Korean trade unions, expressed through civil society consultation structures created under the FTA and supported by the European Trade Union Confederation (ETUC). The European Commission’s report on the results of the FTA up to 2016 concluded that the commercial side was going very well, but that ‘Korea has so far made little progress’ in its commitments under the sustainable development heading (European Commission 2019: 11). Nothing changed in the following years, and so after repeated demands from the ETUC, the EU initiated its dispute procedure. A panel of experts was finally constituted in July 2019, reporting on 25 January 2021. The EU case was based on that rehearsed by trade unions before the ILO. The verdict did not uphold all the EU’s points, but it did conclude, as had the ILO, that a number of laws were incompatible with the fundamental conventions (Novitz 2021). On 20 April 2021 the South Korean parliament voted for the ratification of three ILO fundamental conventions, leaving only No. 105 on the abolition of forced labour for a later date.

The panel’s judgement was helpful as one tool that labour-sympathising Korean politicians could use. Although the EU could impose no further sanctions, opposition was quietened. For example, it was reported in the Korea Herald of 7 July 2020 that the deputy labour minister had warned that approving ILO conventions could head off potential trade risks and pressure from the EU ‘through various measures’. However, in terms of corresponding changes to employment law and improvements in workers’ conditions, the real test is still to come.


Georgia, a former Soviet republic seeking support against a perceived threat from Russia, was much more willing to accept conditions in its trade agreements than Korea. Following the so-called Rose Revolution of 2003, it set out on a course of extreme deregulation, including abolishing its labour inspectorate and approving a new labour code that eliminated much of the previous employment protections. This won praise for a time in the World Bank’s Ease of Doing Business index (World Bank 2006: 3), but its implementation, including summary dismissals of trade union representatives, led to complaints to the ILO in 2008 and 2010 from the Georgian Trade Unions Confederation (GTUC), backed by the ITUC. Subsequent governments partially restored workers’ rights, but with strong opposition from business organisations, including the American Chamber of Commerce. A labour inspectorate was re-established in 2015, but with a limited mandate and powers. Workplace fatalities ran at an exceptionally high level, peaking at 199 in 2018 (ILO 2019a: 39), way above comparable European Union levels, sparking a number of strikes in the mining industry.

One factor that led to greater international acceptability was an Association Agreement with the EU which came into effect in July 2016. This included a detailed timetable for implementing laws to ensure approximate harmonisation with 40 pieces of EU legislation in the areas of employment, social policy and equal opportunities. Change was still a very slow process, though, with new laws only reaching partial compatibility with the EU, and still dependent on overcoming opposition from business organisations which feared the ‘extremely left-wing labour code that operates in highly developed European countries’ (Laliashvili 2009). However, the civil society consultation structures established within the Association Agreement were a useful channel for putting pressure on the Georgian government, which appeared to feel ‘more accountable to the European side than to national organisations’ (Martens et al. 2020: 49).

Decisive change came after political turmoil within Georgia, including international questioning of the country’s civil rights record. A group of MPs, working with ILO advice, put forward a package of labour law reforms which, despite reluctance from business organisations, won unanimous support in parliament on 29 September 2020. Some compromises had been made to keep business on board, but the result was billed as satisfying all international requirements and brought the country close to alignment with the EU.


Vietnam is a one-party state with trade unions subordinated to the political authorities. Trade agreements promise substantial gains in higher exports of footwear and garments into higher-income countries, including the US and EU. Vietnam joined the Comprehensive and Progressive Agreement for Trans- Pacific Partnership (CPTPP) on 12 November 2018 and signed an agreement with the EU on 30 June 2019. Both contain familiar commitments on employment rights and conditions. Much more specific requirements had been set out in a supplementary agreement with the US as part of the Trans-Pacific Partnership (TPP) which preceded the CPTPP and which the then President Trump withdrew the US from in January 2017. This side agreement, described as ‘a complete labor game-changer in Vietnam’ (Trần Thị Kiều Trang and Bales 2017: 74), included a precise timetable for legislative changes before the FTA could take effect that would require membership elections for all union officials and allow for the creation of ‘a grassroots labor union’ within a workplace, with the potential to develop into a national-level union organisation with international links.

Reform was also motivated by chaotic industrial relations in the export- oriented branches of MNCs. Strikes, often spanning workplaces, were a commonoccurrence, motivated by opposition to arbitrary management practices and demands over pay and conditions. Demands were generally quickly conceded, as manufacturers faced tight delivery schedules. These strikes being illegal, strikers established no formal organisations. However, they were a clear embarrassment to a government advertising good industrial relations to investors and also worried about a potential threat to its monopoly of power. Agreement with the US was followed by increasing soul-searching within elite circles in Vietnam over the future of industrial relations.

The pressure was relieved by the US withdrawal from the TPP, a fact that demonstrates in reverse the value of insisting on conditions before signing an agreement. Nevertheless, in January 2021, alongside commitments to complete ratification of the ILO fundamental conventions, a new labour code came into force in Vietnam including provisions for ‘grassroots worker representative organisations.’ However, they are not guaranteed the rights to link up between enterprises and become genuine independent unions. The civil society consultation mechanisms established under the agreement with the EU also face big obstacles as Vietnam’s repressive political system blocks the emergence of an independent civil society. A formal commitment to labour rights may thus be devalued in what remains an authoritarian political system.


Mexico joined with the US and Canada in the North American Free Trade Agreement (NAFTA) in 1994, but subsequent growth in the country was slow and allowed average wages to fall even further behind the US level. Precisely because of its dependence on exports to the US, Mexico was in a weak bargaining position when President Trump promised to renegotiate NAFTA, leading to a new agreement, the United States–Mexico–Canada Agreement (USMCA in the US) which came into force in July 2020. However, Trump’s weakness as he sought re-election also gave an exceptionally strong position to labour in the US, which could press vigorously the issue of Mexico’s industrial relations system, an important factor contributing to low wage levels.

So-called ‘protection contracts’ were signed by union organisations often without input from, or knowledge of, the employees they were representing and often leaving employers with enormous leeway over pay and conditions. Agreements were even signed by incoming multinational companies before they had taken on employees. This system had provoked complaints from 2009 onwards to the ILO, which expressed concern at the contravention of its convention on collective bargaining.

USMCA included four important changes from NAFTA. The first was wording that obviated the need to demonstrate an effect on trade from abuses of labour rights before sanctions could be applied. The second was a requirement for a pay level in parts of motor vehicle production way above the prevailing rate in Mexico. A firm failing to meet the wage threshold would have to pay tariffs even within the USMCA area. The third was an insistence that new laws should be introduced before the agreement would be ratified by the US, ensuring election of union leaders, and secret votes on collective agreements and publication of their contents, including those agreements already signed. The fourth was the Rapid Response Labor Mechanism (RRLM) allowing complaints against individual companies that transgressed workers’ rights with a precise and short timetable for redress. The penalty could be loss of rights to tariff-free exports within the USMCA area.

The third and fourth of these elements were clearly aimed against protection contracts. The changes from NAFTA were acceptable to a new Mexican government which saw benefits in entrenching labour reform in an international treaty, ruling out reversal by a subsequent government. Even business representatives were persuaded that the alternative of no agreement was unthinkable and on 1 May 2019 the Mexican parliament passed unanimously the package of laws aimed at facilitating genuine collective bargaining.

The aftermath has seen the first cases of application of the RRLM to address complaints from both trade unions and US officials, as well as secret ballots on collective agreements and to replace unions that have disappointed their members. It is too early to see the full economic effects and whether manufacturers will raise wages in Mexico or relocate production to the US or to another part of the world. Nevertheless, it seems that a transformation of Mexican industrial relations has been set in motion by USMCA.

Lessons to be learned

Key lessons from these experiences can be summarised under four points:

  1. The strength and effectiveness of labour provisions depend on relative bargaining positions. Georgia, Vietnam and Mexico had more to gain from a commercial agreement and were therefore the most likely to accept meaningful labour provisions. However, trade union persistence also contributed to some gains in Korea, albeit after a slow, long and uncertain process.
  2. The key to change is internal politics. In every case, the (apparently) tough conditions included in the FTA were a useful weapon for labour movements and sympathetic politicians, helping them to promote and cement reforms they had been advocating.
  3. Labour provisions are most effective when they relate to issues confronting the partner country’s labour movement. Requiring ratification of ILO conventions could have appeared abstract, but it was linked to countering concrete abuses of labour rights in Korea. In other cases, much more specific requirements for change were written into the agreement.
  4. There is no question that sanctions work. Even the weakest, a condemnation as applied to Korea, can help strengthen the position of politicians pressing for change. The best time to apply sanctions is still before an agreement takes effect, as shown in the cases of Vietnam and Mexico, with the possible threat of a quick application of sanctions also appearing to be making a difference in Mexico.

References - see PDF

Table of contents

Making labour provisions in free trade agreements work