Servicification of GVCs through deep service provisions: Uncovering new insights from structural gravity and machine learning

In recent decades, there has been a notable increase in linkages of services in decoupling global value chains (GVCs) and a surge in regulatory mechanisms embedded in service provisions in trade agreements. Existing literature tried to empirically link the impact of such service provisions on GVC-related services, but none focused on identifying relevant service provisions. This study is a novel attempt in this direction using a machine learning algorithm augmented in gravity modelling. Building on the identified service provisions, the study quantifies their impact on GVC-related services conditioned on the countries' income levels. The study also conducts the general equilibrium analysis by simulating a scenario incorporating identified service provisions in the India-ASEAN trade agreement. The analysis finds that few service provisions exist that enhance the share of foreign service inputs in manufacturing exports of the countries involved in GVC-related service participation. Moreover, the impact is heterogeneous regarding benefits to the developing countries as a destination of service-value added. Finally, the study shows that introducing selected service provisions in existing trade agreements can potentially increase welfare and service trade.