One implication of the pollution haven hypothesis is that countries export more by applying more lenient environmental regulations. Most studies that apply gravity-type equations do not find robust support for environmental regulations to affect bilateral exports. In this paper, we show that one can obtain robust negative effects of stringency, as long as gravity equations are well specified with respect to theory. Our results, based on the European data, are both very consistent with US studies on environmental regulations and another line of very recent studies that infer non-biased price or substitution elasticities from trade equations. We show that more stringent environmental regulations, when depicting a pure cost effect, are reducing exports. The coefficient is even larger in the case where exporting countries are Central and Eastern European countries, comparing to the EU15. Further, we show that there is no significant difference in the impact of regulations on trade in case of dirty and clean sectors. Finally, when using GMM estimation, our environmental stringency coefficient gets significantly reinforced. Copyright Blackwell Publishing Ltd 2005.

Recent economic geography and trade empirical studies based on monopolistic competition suggest high levels of trade price elasticities~between 3 and 11!. However, price elasticity estimations in trade equations using unit values as price proxies usually lead to lower values of around unity. We show that those inconclusive results may be due to some misspecification in these equations as well as measurement errors in prices. When suitable instrumental variables are used, within a panel of industrialized countries, we obtain high price elasticities, the majority ranging from 1 to 13. The highest estimates correspond to industries producing homogeneous goods. JEL classification: C2, C3 and F1.Les élasticités prix des importations : un nouveau coup d'oeil aux résultats. Plusieurs études empiriques récentes fondées sur des modèles de concurrence monopolistique exhibent des élasticités prix estimées des échanges élevées~entre 3 et 11!. Or, la plupart des élasticités prix estimées dans la littérature approchant les prix par des valeur unitaires sont plus faibles, de l'ordre de 1. Nous montrons que ces résultats non concluants pourraient provenir d'une mauvaise spécification des équations d'échanges ou d'erreurs de mesure sur les prix. Quand ceux-ci sont correctement instrumentés, sur un panel de pays industrialisés, nous obtenons des élasticités prix élevées~de 1 à 13!. Les plus fortes correspondent aux secteurs produisant des biens homogènes. This paper benefited from the financial support of both the Commissariat Général au Plañ N.4-98! and the French Ministry of the Economy, Finance and Industry. We would like to thank especially Thierry Mayer, since he kindly provided us with a transport cost proxy~weighted distance! for the ten European countries included in our sample. Special acknowledgment is also addressed to James Rauch, who provided us with the data relative to his industry classification in Rauch~1999!. Moreover, we thank participants at the World Econometric Society Conferencẽ Seattle, 2000! and the University of Paris 1 Panthéon -Sorbonne Workshop, especially Lionel Fontagné, Mathieu Crozet, and François Gardes, for helpful comments. We are very grateful to Françoise Le Gallo and Jean-Marie Lozachmeur for research assistance. Finally, we thank an anonymous referee for his or her valuable comments and suggestions.