The gravity of intermediate goods

Abstract

One of the puzzles of the gravity literature is the persistent effect of distance on trade flows, despite the dramatic fall in trade costs during the last few decades (Disdier and Head, 2008). A possible reason for the \distance puzzle" is that trade in intermediate goods, which has risen dramatically during this period due to the emergence of global value chains, may be more sensitive to distance than trade in final goods. Using a dataset of bilateral import flows covering 5000 products and more than 200 countries over the 1998-2011 period, we show that intermediate goods are indeed more sensitive to distance than final goods, with differentiated inputs exhibiting the highest distance elasticity. The results are robust to including different sets of controls, and using different samples and econometric methodologies. They suggest that sourcing inputs from nearby countries helps final good producers to better coordinate with their suppliers, monitor their production, and insure the timely delivery of inputs that need to be tailored to their needs.