Many studies consider the potential for preferential trade agreements (PTAs) to have differing effects on trade. Kohl (2014) and Baier et al. (2019) show that some PTA’s promote trade while the majority of PTAs have no significant effect. Some even lower trade. Why do these differing cases arise? One possibility is that the effects of trade agreements depend on specific provisions – provisions that differ across agreements. Another possibility is the potential for PTAs to impact trade differently depending on the presence of certain bilateral characteristics between trading nations such as physical distance or metaphorical types of distance such as culture or language. In my dissertation, “Heterogeneous Effects of Trade Agreements on Trade,” we explore these two avenues separately.In the first chapter we consider if differences in the prevalence of corruption between members of a PTA make trade agreements more or less effective at boosting trade. Such differences could create more uncertainty that limits the potential for trade even if a trade agreement lowers barriers, implying that such agreements will not boost trade. On the other hand, trade agreements could be most effective in such disparate countries. Not only might trade agreements remove barriers used by corrupt officials to extort firms, but a trade agreement could reduce the uncertainty of operating in a different business environment by establishing rules and regulations. Results in this paper are allowed to differ across several dimensions, including extensive versus intensive margin, whether the exporter or importer is more corrupt, and between South-South and South-North trade. Using a gravity model of trade spanning a panel of countries from 1996 to 2017, we find that PTAs increase trade more along the intensive margin when importing countries are more corrupt but boost trade more along the extensive margin when exporting countries are more corrupt. Results are stronger for trade between South-South (S-S) countries than between North-South (N-S) countries. Chapter two examines how specific provisions within trade agreements – particularly, provisions regarding environmental standards – affect trade between members and non-members. While there is a rising trend to incorporate different types of environmental provisions in preferential trade agreements (PTAs), few studies took explicit steps to assess the trade consequences of environmental provisions in PTAs. This paper employs a gravity model over the period from 1984 to 2016 and uses a new detailed dataset on a broad range of environmental provisions in PTAs to fill the gap in the literature by looking at possible trade diversion effects from trade agreements with deep environmental clauses. We follow Mattoo et al. (2017) and construct an index that captures importers’ average depth of trade agreements with the rest of the world where depth is taken as the extent that environmental provisions are covered. The inclusion of this depth variable allows us to see if any trade diversion effect arises from trade agreements with deep environmental provisions. We specifically focus on exporters with low environmental standards, as those are the countries that are likely to “host” trade in environmentally unsustainable goods. We also differentiate between different types of environmental policies and concentrate on trade in “dirty” products. Our results suggest that environmental provisions in PTAs are an effective tool of promoting environmentally sustainable trade in the world, as these types of policies tend to reduce “dirty” trade even with non-member nations. Finally, the third chapter considers the heterogeneous design of PTA’s more broadly, looking at the trade effects of different policy areas within trade agreements, while differentiating their impact on trade in new product varieties of goods versus trade in existing products. We specifically focus on 18 “core” provisions that Hofmann et al. (2019) mark as most economically relevant policies. We further distinguish three types of policies within the “core” group of provisions, namely: i) provisions that directly liberalize trade through either reduction in tariffs or simplification of standards, ii) policies that enable signatory nations to compete on equal grounds, and iii) provisions that specify the rules of investment. Previous studies that consider the effects of trade agreements on the margins of trade have either focused on the effects of different types of PTAs, rather than specific policies, or used limited data and outdated methodologies. We are contributing to the literature by assessing the impact of different groups of policies on the margins of international trade using a highly disaggregated dataset covering a large number of countries and years. We also employ Factor Analysis to check robustness of our findings using regular count indices. Our results indicate that provisions that tend to reduce barriers to trade through either simplification of standards or reduction in monetary charges tend to increase trade in existing varieties of goods, while the effect of investment provisions is either insignificant or might actually lower trade.

O debate atual sobre as consequências da proliferação de Acordos Comerciais Regionais surge no meio de uma crise de governança no comércio internacional, o qual também questiona o papel da OMC como Organização Internacional que governa este cenário através de princípios e normas multilaterais. No entanto, cientes desta situação, os Estados parecem manter esforços para lidar com tais problemas através da negociação de mecanismos multilaterais para melhorar a governança do sistema multilateral. O Comitê de Acordos Comerciais Regionais (CRTA) que existe há mais de duas décadas é uma peça-chave para compreender as razões pelas quais tais esforços parecem infrutíferos, considerando o fato de que se trata do órgão multilateral encarregado do controle de Acordos Regionais. Assim, a presente pesquisa procura analisar o trabalho e evoluções desse Comitê, considerando diferentes fatores que incidem em seu desempenho, e determinar se um trabalho como o feito pelo CRTA poderia ser uma eficiente forma de governar as relações que surgem da atual organização da produção internacional. / The current debate about the consequences of the proliferation of Regional Trade Agreements arises in the middle of a governance crisis in international trade, which has also put into question the role of the WTO as an International Organization governing this scenario through multilateral principles and rules. However, aware of this situation, States seem to keep making efforts to deal with these problems through the negotiation of multilateral mechanisms to enhance the governance of the international trade system. The Committee on Regional Trade Agreements (CRTA), existing for more than two decades, is a key piece to understand why these efforts seem fruitless. This research attempts to analyze its work and evolutions, considering different factors that have incidence in its performance, and to determine if it is possible that a work such as the performed by the CRTA could be an efficient way to govern the relations arising from the current organization of international production.

Dans un contexte où les négociations commerciales multilatérales languissent dans une impasse, les accords commerciaux régionaux ACR prennent de l’élan. Cette thèse doctorale cherche à faire avancer la connaissance sur ce domaine. C’est grâce au modèle de gravité du commerce que trois chapitres supportés par des analyses économétriques appliqués ont été mis au point.Le premier chapitre examine les effets sur les flux bilatéraux de commerce attribuables aux ACR, le système généralisé de préférences (SGP) et l’appartenance à l’Organisation mondiale du commerce OMC. Plusieurs spécifications économétriques et techniques d’estimation ont été testées. Particulièrement Poisson Pseudo Maximum Likelihood (PPML), qui se présente comme la technique la plus recommandée pour contenir des biais bien connues et des problèmes d’endogénéité. Cette recherche a été conduite avec un modèle de gravité du commerce international qui comporte 153 pays sur la période 1980-2012. Les résultats montrent systématiquement qu’un effet positif et significatif sur les flux bilatéraux de commerce est à attendre après l’entrée en vigueur d’un ACR. De même, des effets positifs mais peu importants, voir nuls sont accordés à la participation au sein de l’OMC. La spécification qui utilise PPML et qui contrôle l’influence de l’hétérogénéité inobservable montre un effet non-significatif pour le SGP. Le deuxième chapitre, coécrit avec mon directeur de thèse Jean-Marc Siroën, explore quel est l’effet de l’hétérogénéité des ACR sur le cadre de l’approfondissement de l’intégration. Nous envisageons pouvoir déceler si les ACR qui sont plus profonds contribuent plus à la création de commerce que ceux qui sont moins profonds. Nous avons recours à deux bases de données récemment ouverts au public. La première appartenant à l’OMC et la deuxième a la World Trade Institute (WTI-DESTA). Nous procédons à créer des indicateurs crédibles de l’approfondissement de l’intégration pour passer à les tester dans un modèle de gravité. Nous trouvons qu’un effet positif et significatif peut-être accordé aux accords les plus profonds, indépendamment que l’indicateur testé soit un indicateur additive où un indicateur obtenu par l’Analyse de correspondance multiple (ACM). De même cet effet est constaté pas seulement dans les accords qui comportent des clauses classiquement négociées sur le cadre de l’OMC, mais aussi dans les accords qui dépassent cette dimension. Le troisième chapitre se consacre à étudier l’existence des potentiels de commerce entre la Colombie et l’Union Européenne. Des prédictions dans l’échantillon après des estimations avec PPML et effets fixes qui varient dans le temps nous indiquent que des potentiels de commerce existent avec l’Autriche, la République Tchèque, la Finlande, la France, l’Allemagne, la Hongrie, la Suède et la Pologne. Dans le sens inverse la Suède, l’Irlande, la Finlande et Pologne détiennent une marge importante à gagner dans le marché colombien. Des tests de sensibilité ont été effectués pour garantir la robustesse de ces résultats. / Regional trade agreements (RTAs) have surged in a context of stalled multilateral trade negotiations. This doctoral thesis intends to advance scientific knowledge in the field. Thus, thanks to a gravity model theoretical framework, three chapters of applied empirical econometrics analysis have been completed. The first chapter examines the effects of RTAs, the Generalized System of Preference (GSP) and World Trade Organization memberships on bilateral trade flows. I put into practice different econometric specifications and estimation methods, notably Poisson Pseudo Maximum Likelihood (PPML), which is the one that better seems to contend with well-known biases and endogeneity problems. I conduct this research with an international trade gravity model estimated across 153 countries from the year 1980 to 2012.I consistently found a strong positive impact of regional trade agreement RTAs on most specifications and low or non-significant results for WTO membership. The estimates from the PPML method that includes controls for unobserved heterogeneity show non-significant effects of the Generalized System of Preference (GSP) on trade.The second chapter, co-authored with my supervisor Jean-Marc Siroën, explores the effect of heterogeneity of RTAs in the scope of deep integration. We intend to determine if deeper RTAs promote trade more effectively than less ambitious agreements. We make use of two recently available data sets from the World Trade Organization (WTO) and the World Trade Institute (WTI-DESTA) to generate credible indicators of deep integration. Additive and Multiple Correspondence Analysis derived indicators for the depth of the agreements are then computed and their significance is tested in a gravity model. We find that deeper agreements increase trade more than shallow ones, whereas the provisions they included are within or outside of the WTO domain.The third chapter investigates the existence of trade potentials between Colombia and the EU. I obtain in-sample predictions after the estimation of a gravity model with the Poisson Pseudo Maximum Likelihood estimator. I control for unobserved omitted variable bias by the inclusion of exporter and importer time varying fixed effects, and run a series of sensitivity analysis.Untapped trade potentials are found between Colombia and a group of EU countries in both directions of the trade flows. Exports from Colombia have a gap to bridge with Austria, Czech Republic, Finland, France, Germany, Hungary, Poland and Sweden. In the other direction, Sweden, Ireland, Finland and Poland have an interesting margin to gain in the Colombian market.