Breaking badly: The currency union effect on trade

•In this paper, we revisit the important policy question of whether CUs increase trade using standard tools from modern applied microeconomics.•Approaches adopted in this paper include plotting pre- and posttreatment trends, adopting more-suitable control groups, controlling for other major geopolitical factors, controlling for dynamics, estimation using PPML, and adopting a synthetic control approach.•We find estimates of CUs on trade that are usually, but not always, positive—and noisy—as they are often not statistically significant and sensitive to the specification.•When using OLS with controls, a synthetic control approach, or a high-dimensional PPML estimator, we do not find evidence that CUs have a large positive impact on trade. As several European countries debate entering, or exiting, the euro, a key policy question concerns how much currency unions (CUs) affect trade. Despite the longstanding academic debate on the topic, recent research has continued to find that CUs exert a large effect on trade. We find, by contrast, that the sizeable recent estimated impact of CUs on trade is driven by other major geopolitical events and is also sensitive to dynamic controls. Overall, using various specifications and controls, we estimate that the impact of CUs on trade is indistinct from zero but with relatively large standard errors.