In recent years, the content of preferential trade agreements (PTAs) has expanded considerably in terms of both the number of issues that are covered and the detail which these issues are dealt with. Other than the standard chapters in preferential trade agreements dedicated to removal of tariff and non-tariff barriers, and the regulation of cross-border issues related to investment, competition, services, and intellectual property rights, PTAs increasingly feature provisions on civil and political rights (CPR), economic and social rights (ESR), and environmental protection (EP). These latter provisions are not directly related to trade (henceforth non-trade issues, or NTI), yet they might affect firms engaged in trade and investment operations across the border.
The relation between free trade agreements (FTAs) and FDI is not straightforward due to a variety of motives for investing abroad. Trade and investment can be either complements, in case of input-seeking vertical FDI, or substitutes, in case of market-seeking horizontal FDI, with FTAs likely to stimulate the former (e.g. Kox and Rojas‐Romagosa 2020, Osnago et al. 2020) and deter the latter (e.g. Bergstrand and Egger 2007). An even more complex relation can emerge when FDI is characterised by mixed motives (Baldwin and Okubo 2014).
Non-trade related provisions in PTAs
The effect of non-trade issues on FDI is so far unexplored. There are a variety of possible effects. For instance, clauses restricting pollution or imposing a minimum age at work could hinder MNEs’ entry into certain markets due to increased operational costs. On the other side, an improvement in civil and political rights might reflect progress on governance and strength of institutions, which can increase the incentives to invest. Finally, there could also be no effect if the function of NTIs is to prevent the worsening of certain minimum standards.
We investigate the FDI-NTI relationship in search of empirical regularities that could shed light on how NTIs in PTAs affected greenfield bilateral FDI (Di Ubaldo and Gasiorek 2020). As most PTAs contain some non-trade related provisions, it is important to distinguish the extent to which NTIs are covered in a PTA. To do so, we exploit data from Lechner (2016) on the degree of legalisation of CPR, ESR, and EP in PTAs. The legalisation score goes beyond the mere identification of a provision and captures, for instance, whether an anti-air pollution clause is included in the main text of the agreement (level of obligation), whether the clause refers to an international treaty (level of precision), and whether the parties intend to consult international organisations or NGOs (level of delegation).1 A PTA’s legalisation score is then obtained by aggregating the three dimensions of obligation, precision, and delegation in each area of CPR, ESR and EP.
Figure 1 Evolution of non-trade issues in PTAs by type
Source: Authors’ elaboration on Lechner (2016).
Figure 1 shows the average NTI legalisation scores of PTAs which entered into force within 5-year intervals, classified into CPR, ESR, and EP categories. All NTIs saw an increase in legalisation over time, in particular since the mid-1990s. The economic-related NTIs, ESR and EP provisions, saw a faster growth in legalisation than CPR provisions, possibly signalling a growing tendency of exploiting PTAs to establish a level-playing field among trade partners.
To assess how the NTIs affect cross-border investments, we extract data from fDi Markets on 71,572 greenfield investment projects between countries partners of 285 PTAs in force between 2003 and 2017.2 The mean NTI legalisation is substantially higher for agreements concerning FDI between high-income (North) countries, compared to PTAs concerning FDI between middle and low-income (South) countries, with the largest gap emerging for ESR and EP provisions (Table 1). Interestingly, the mean score for the subsample of ‘mixed’ North-South FDI is close to that of North-North FDI, suggesting that it might be the high-income countries that demand the inclusion of NTIs in their agreements.
Table 1 Mean legalisation scores of PTAs by FDI sub-groups
Note: We separate the countries in our data into North countries (high-income) and South countries (middle- and low-income) according to the World Bank classification. Note that PTAs can belong to more than one of these subgroups, as some agreements include groups of countries with heterogenous income. This is evident from the figures at the bottom of Table 1, as the sum across columns 2-4 exceeds the total in column 1. Source: authors’ elaboration on Lechner (2016) and fDi Markets data.
How does FDI respond to NTI provisions?
We exploit a structural gravity setting to estimate how an increase in NTI legalisation of PTAs affects the flow of greenfield FDI. Across a wide range of exercises,3 the recurring result appears to be that a higher degree of NTI legalisation leads to a lower flow of bilateral FDI. The effects are comparable in magnitude across the three types of provisions, although they are more robust and statistically strongest for CPR, and weakest for ESR provisions.
We estimate an elasticity ranging from -0.16 (CPR provisions) to -0.08 (EP provisions) in the aggregate sample.4 When distinguishing the NTI effect depending on the destination and origin country of the investment, we find that the negative effect emerges in a robust way for FDI directed South, and between South-South countries in particular: on this latter subsample, the FDI elasticity with respect to NTI increases to about -0.27. The effect of NTIs on FDI is again found to be negative when restricting the estimation sample to FDI in manufacturing: the effect is again largest for South-South FDI for CPR and EP provisions, with an elasticity of about -0.33.
Discussion and conclusion
Finding a unique argument which can explain these negative effects is challenging. The design of NTIs in PTAs, in fact, differs not only over time and across countries, but also within countries across partners. For instance, the EU-CARIFORUM and the EU-Cote d’Ivoire agreements, both signed in 2008, have a legalisation score of 36 and 18, respectively. The heterogeneity in the NTI design extends to the different types of provisions: the EU tends to demand the inclusion of more stringent CPR provisions in its agreements with small neighbouring countries such as Moldova and Georgia, while it tends to negotiate a larger number of economic-related provisions with large trade partners such as Canada, South Korea, and Japan (Borchert et al. 2020a).
One likely explanation is that NTIs deter FDI because they impose higher costs to the operations of multinationals, and hence higher NTI’s may lead to investment diversion. This argument applies rather easily to ESR and EP provisions, which are likely to increase labour costs and impose adherence to production standards which could make the destination country a less attractive location for FDI. CPR provisions, on the other side, are the most enforceable among the NTI clauses (Borchert et al. 2020b), and could deter FDI because a violation can lead to sanctions (up to the suspension of the PTA in EU agreements), with likely repercussions for firms involved in cross-border operations.
Baldwin, R and T Okubo (2014), “Networked FDI: Sales and sourcing patterns of Japanese foreign affiliates”, The World Economy 37(8): 1051-1080.
Bergstrand, J H and P Egger (2007), “A knowledge-and-physical-capital model of international trade flows, foreign direct investment, and multinational enterprises”, Journal of International Economics 73(2): 278-308.
Borchert, I, P Conconi, M Di Ubaldo and C Herghelegiu (2020a), “The Pursuit of Non-Trade Policy Objectives in EU Trade Policy”, CEPR Discussion Paper No. 14655.
Borchert I, P Conconi, M Di Ubaldo and C Herghelegiu (2020b), “EU trade policies: Carrot-and-stick mechanisms in the pursuit of non-trade policy objectives?”, VoxEU.org, 5 May.
Di Ubaldo, M and M Gasiorek (2020), “Non-trade provisions in trade agreements and FDI”, WPS 2120, Department of Economics, University of Sussex Business School.
Dür, A, L Baccini and M Elsig (2014), “The design of international trade agreements: Introducing a new dataset”, The Review of International Organizations 9(3): 353-375.
Kox, H L and H Rojas‐Romagosa (2020), “How trade and investment agreements affect bilateral foreign direct investment: Results from a structural gravity model”, The World Economy.
Lechner, L (2016), “The domestic battle over the design of non-trade issues in preferential trade agreements”, Review of International Political Economy 23(5): 840-871.
Osnago, A, N Rocha and M Ruta (2019), “Deep trade agreements and vertical FDI: The devil is in the details”, Canadian Journal of Economics 52(4): 1558-1599.
1 Lechner’s (2016) data structure is very rich: she codes 262 data points, then aggregated in 126 variables, referring to various CPR, ESR and EP clauses that can be contained in FTAs and their degree of obligation, precision, and delegation.
2 We restrict our sample to country-pairs members of PTAs and assess the effect of NTIs exploiting the variation in the continuous legalisation score across PTAs, and over time.
3 The results are robust to aggregating the FDI data over two- and three-year intervals, and to exploring contemporaneous and lagged NTI variables.
4 This effect is robust to controlling for the presence of bilateral investment treaties, EU membership, the depth of the PTA (DESTA depth index; see Dür et al. 2014) and the presence of substantive investment-related provisions in the PTA.