Points clés :

• This paper offers an estimate of the firm level price elasticity of exports using an original instrumental variable strategy. Our results point robustly to an estimate around 5.
• Our results show that the international elasticity puzzle is worse than previously thought. Not only is the elasticity of exports higher for tariffs than for exchange rates, the elasticity of exports to export prices is much larger than those two.
• We show that an estimate of elasticities of exports to exchange rates and tariffs that does not take into account the endogenous reaction of export prices is a mix of two opposite effects: the elasticity of substitution between home and foreign goods and the elasticity of exports to the endogenous reaction of export prices to the exchange rate or tariff shock.

 Résumé :
We estimate three international price elasticities using exporters data: the elasticity of firm exports to export price, tariff and real exchange rate shocks. In standard trade and international macroeconomics models these three elasticities should be equal. We find that this is far from being the case. We use French firm level electricity costs to instrument for export prices and provide a first estimate of the elasticity of firm-level exports to export prices. The elasticity of exports is highest, around 5, for export prices followed by tariffs, around 2, and is lowest for the real exchange rate, around 0.6. The large discrepancy between these elasticities makes us conclude that the international elasticity puzzle is actually worse than previously thought. Moreover, we show that because exporters absorb part of tariffs and exchange rate movements, estimates of export elasticities that do not take into account export prices are biased.

 Mots-clés : Elasticity | International Trade and Macroeconomics | Export Price | Firm exports

 JEL : F14, F18, Q56