Published online by Cambridge University Press: 12 March 2018
We study the effect of exporting on markups building on two stylized facts: (1) exporters charge higher markups than nonexporters and (2) firms increase markups when they start to export. These facts suggest that exporting increases markups, but the causal relationship has not been studied directly. To do so, we modify Melitz and Ottaviano (2008) by adding decreasing returns technologies and demand and productivity shocks to account for sales correlations across markets. We calibrate and simulate a trade cost reduction. Old exporters increase markups on average, while new ones reduce them. Three mechanisms matter: (1) cost reductions are not fully passed on to prices, (2) firms expand, increasing marginal cost, and (3) foreign demand is more elastic than domestic demand. The first effect dominates along the intensive margin, while the others prevail along the extensive margin. Thus, exporters charge larger markups in spite of exporting, not because of it.
We thank Bernardo Blum, Amit Khandelwal, Giammario Impullitti, Sam Kortum, Peter Schott, and two anonymous referees for helpful comments. We benefited from talks at the Kansas Fed, ILADES, NASME 2014, SED 2014, WEAI 2014, and RIDGE 2014. Rubini gratefully acknowledges financial support from FONDECYT (Grant 11140147).