Building BLOCS and stepping stones: Combined data for international economic and policy analysis

The question of whether trade agreements are “stepping stones” or “stumbling blocks” to multilateral trade liberalization is not a new one; however, the empirical methods and the quality of the data used to address this question are continually improving. Our paper explores this familiar question using a robust combined dataset and advances in structural gravity analysis to offer insight into regional integration agreements and trade networks. We introduce the Bilateral Longitudinal Observations and Country Statistics (BLOCS)Footnote1 database and demonstrate variation in Poisson Pseudo Maximum Likelihood (PPML) estimations using measures from varying sources. The data includes observations between 218 sovereign states, and their trading partners, over 60 years (1963–2022). We employ four measures of trade flows, and two measures for trade agreements, accounting for varying definitions and reporting practices. Observations also include information on agreement depth and country attributes to contextualize existing literature and further understand the relationship between international trade and agreement provisions. Differences observed between measures of Regional Trade Agreement (RTA) pairs indicate that the methodology for coding trade agreement participation matters. Our findings also suggest that variations in agreement details predict variation in total trade, thus supporting the hypothesis that not all trade agreements are created equal. We find that it is the institutional details that determine whether a trade agreement will be a “stepping stone” or a “stumbling block” to multilateral trade liberalization.