Endogeneity of currency areas and trade blocs: evidence from a natural experiment

This paper draws on a natural experiment to examine the effects of policy arrangements on international trade. We study data on trade and currency bloc formation in Europe after the Great Depression. Far removed from being customs or currency unions, these blocs could not create much trade and should be mere placebos. Yet under conventional approaches to the gravity equation, they exhibit highly significant and sometimes very large trade effects. We employ treatment effect methods from labor econometrics to identify endogeneity both along the time axis and in the cross section. We find pervasive evidence of such endogeneity, which standard estimates of the gravity equation fail to detect. These findings are confirmed by matching models designed to eliminate the endogeneity of bloc formation itself. Our results caution against the significant and high trade creation effects of political arrangements often reported in the gravity literature.