Trade agreements increasingly include a focus on sustainability goals. This column examines the impact of environmental, labour, and civil rights provisions in trade agreements on related outcomes in developing countries. Enforceable provisions have little effect on performance indicators, whereas non-enforceable provisions in non-EU preferential trade agreements are accompanied by a deterioration in several outcome measures. Moreover, countries that have accepted enforceable non-trade provisions obtained less development aid, whereas non-enforceable environmental provisions are accompanied by more EU aid. These findings suggest ‘soft law’ non-trade commitments are more likely to be accompanied by financial and technical support.
Over the last few decades, trade agreements have increasingly been seen as a potential tool to address environmental and social issues. As climate change intensifies and human rights violations remain persistent in many parts of the world, the question of whether trade deals can be an effective mechanism for attaining sustainability goals is an important one. This issue is particularly salient in the EU, where the promotion of core values is embedded in its founding treaties. The EU has positioned itself as a leader in incorporating non-trade provisions (NTPs) into trade agreements, but the trend is a general one. Almost two-thirds of all preferential trade agreements (PTAs) in force include provisions that go beyond traditional trade liberalisation commitments (see Figure 1), addressing objectives as diverse as environmental protection, labour standards, and civil and human rights (Hofmann et al. 2019).
Figure 1 Evolution of provisions in preferential trade agreements
Source: Hofmann et al. (2019)
Existing research has focused primarily on the implications of non-trade provisions for trade (e.g. Winters 2023, Hoekman et al. 2023). Limited attention has been given to their impact on associated non-trade outcomes (Fernandes et al. 2023). While the motivation for such provisions is to improve standards in developing countries, a key question is whether they are effective in doing so.
In Francois et al. (2025), we investigate whether non-trade provisions in trade agreements addressing environmental protection, labour market regulation, and civil rights promotion affect the performance of developing countries with respect to related non-trade outcome indicators. The analysis is motivated by – and contributes to – two dimensions of the trade and development policy debate.
The first debate concerns the design of non-trade provisions in preferential trade agreements, specifically whether provisions should be enforceable through legal action, allowing trade to be used as a sanctioning device. Advocates of enforceable provisions argue that this is necessary for non-trade provisions to be effectively implemented and that soft law or best efforts-type non-trade provisions will not lead to meaningful improvements in desired non-trade outcomes (Bronckers and Gruni 2021). In contrast, development professionals often argue in favour of soft law provisions, based on the view that these are more likely to improve outcomes of interest by providing a focal point for dialogue and cooperation between preferential trade agreement signatory governments and engagement with stakeholders (Fiorini et al. 2019).
The second policy debate concerns the role of development aid to strengthen the capacity of developing countries to implement non-trade provisions and achieve associated sustainable development goals. Many development professionals argue that technical and financial assistance targeting specific non-trade goals is needed to complement non-trade provisions, whether enforceable or not (Berliner et al. 2015). We expect non-enforceable non-trade provisions to be conducive to more development assistance, whereas enforceable non-trade provisions may not lead to more aid. Insofar as signatory countries have put in place the associated regulatory regime or legislation (i.e. have implemented the provisions), this could obviate the need for assistance. Moreover, if dispute settlement procedures are regarded as an effective means to ensure compliance, this may lead to less aid allocated to areas covered by enforceable non-trade provisions.
Non-trade provisions do not lead to improvement in most outcome indicators
In Figures 2 and 3 we report estimates of the impact of enforceable and non-enforceable environmental, labour, and civil rights provisions on corresponding outcomes. We use a synthetic difference-in-differences methodology, comparing countries that signed agreements with provisions to those that signed similar agreements without the provisions of interest. In both figures, a red marker indicates a significant deterioration in the related non-trade outcome, while a green one indicates a significant improvement (with grey colour indicating estimates that are not statistically significant).
Looking at non-enforceable provisions, we only find a significant improvement in one environmental indicator, PM2.5, in the case of the agreements signed with the EU. Nonetheless, the change appears rather small, amounting to only a 5% improvement compared to the sample mean. Conversely, non-EU agreements are more likely to induce a deterioration in some of the environmental outcomes considered, presumably reflecting the consequences of an increase in economic activity following the entry into force of a preferential trade agreement. We find no evidence of any impact of non-enforceable provisions pertaining to labour market regulation and standards or civil rights in signatory countries.
In the case of enforceable provisions, we do not find any significant relationships, with the exception of a deterioration in one of the labour market variables, workers’ representation, in the case of non-EU agreements.
Figure 2 Non-enforceable provisions vs. no non-trade provisions (standardised average treatment effect)
Overall, the results reveal few statistically significant effects, and for almost all outcome variables where significant changes are found, the relationship is negative, with non-enforceable non-trade provisions in non-EU preferential trade agreements associated with a worsening of some environmental outcome indicators. This is especially the case with emissions.
Non-trade provisions may be more effective if flanked by financial support
The second policy debate motivating our analysis concerns the relationship between inclusion of non-trade provisions and the allocation of official development assistance (ODA). We assess whether policy-specific (sectoral) ODA varies with non-trade provisions covering the sectors (policy areas) concerned, and whether the overall level of ODA is affected by different types of provisions and their enforceability. This question is relevant for governments in developing countries that may accord higher priority to policy areas and projects not covered by non-trade provisions and thus seek additional aid to pursue these as part of the negotiation of a deep preferential trade agreement. Successful issue linkage strategies by developing countries might be reflected in increased overall official development assistance allocations.
Figure 3 Legally enforceable provisions vs. no provisions (standardised average treatment effect)
We find non-enforceable environmental provisions tend to correlate with increases in official development assistance, but only in EU trade agreements. Enforceable provisions show no significant relationship with aid disbursement. Furthermore, we find a negative association between the acceptance of enforceable non-trade provisions and official development assistance targeting the conditions in the labour market or the promotion of civil rights. This finding suggests different mechanisms may be at work when countries incorporate environmental and social objectives into their trade relationships. While enforceability – with the implied commitment to implement a non-trade provision – appears to be a substitute for development assistance for non-EU advanced economies, the results suggest official development assistance is a complementary policy instrument for the EU.
Figure 4 Official development assistance and non-trade provisions in preferential trade agreements
Overall, our results suggest that non-trade provisions in trade agreements have little impact on the associated policy outcomes. While some effects are observed for environmental performance indicators, particularly in non-EU preferential trade agreements, where there is evidence of deteriorating environmental outcomes, no significant effects are found for labour or civil rights provisions. Interestingly, observed significant relationships mostly involve non-enforceable environmental provisions, with enforceability appearing to prevent the deterioration of the associated non-trade outcome indicators in developing countries. Finally, non-trade provisions are not consistently linked to an increase in development assistance. While some complementary relationships exist for non-enforceable environmental provisions in EU preferential trade agreements, enforceable labour provisions in EU agreements and civil rights provisions in non-EU agreements actually may substitute for aid, highlighting differing priorities among high-income countries.
Source : VOXeu
