Working Paper: CEPR ID: DP14123
Authors: William Connell; Emmanuel Dhyne; Hylke Vandenbussche
Abstract: This paper uses Business to Business (B2B) transaction level data. It shows that manufacturing firms that initially export via a wholesaler are much more likely to become direct exporters to the same destination in subsequent periods. Theoretically, we rationalize this finding by demonstrating how a connection to a wholesaler reduces uncertainty about the foreign demand. In the data we isolate the channel for demand learning from productivity spillovers. Non-exporting manufacturing firms, previously serving a foreign destination through an exporting wholesaler, have a much higher probability of becoming direct exporters to the same export market in subsequent periods. A connection to an exporting wholesaler results in a probability of exporting to the same destination that is six times higher than a comparable firm without any exposure to the foreign destination.
Keywords: B2B data; learning about demand; direct and indirect exports
JEL Codes: F14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Connection to wholesalers (L81) | Reduces uncertainty about foreign demand (D89) |
| Learning about foreign demand (F10) | Increases expected profitability of exporting (F10) |
| Connection to wholesalers (L81) | Transition from indirect to direct exporting (F10) |
| Non-exporting firms connected to exporting wholesalers (F10) | Direct exporting to the same destination (F10) |