Trade agreements and the design of democracy-related provisions (TRADEM): The creation of a new data set

Abstract Scholars have long sought to understand the relationship between trade agreements and democracy. This research note contributes to studying this relationship by presenting a new, comprehensive database on democracy-related provisions in preferential trade agreements (PTAs). The data is unique in its coverage and data creation approach. First, it is the most systematic and in-depth collection of democracy-related provisions in PTAs. It includes obligations to promote democracy, the protection of individual rights, alongside other important democracy-related provisions to increase transparency and stakeholder participation, and to preserve policy space. Second, it combines manual text coding and machine learning approaches to generate the dataset and two types of new indices. The research note first provides an overview of the existing literature on the PTA-democracy nexus. Then, it outlines our conceptualization of how particular PTA provisions are relevant from the viewpoint of democracy. This leads to a new typology covering six dimensions encompassing seventy-two democracy-related provisions. The research note then describes how we combine manual coding and machine learning to create the TRADEM database. We further present descriptive statistics focusing on categories and trends over time and region for a total of 792 PTAs signed between 1949 and 2022. We conclude by outlining potential applications and research avenues.