Search

Does leaving a currency union reduce international trade? This paper reexamines time series estimates of currency unions on trade from a historical perspective using a dynamic gravity equation and by conducting in-depth case studies of currency union breakups. The early large estimates are sensitive to dynamic specifications, and were driven by omitted variables, as many breakups were caused by warfare, communist takeovers, coup d'etats and other major geopolitical events. The methodology has general applicability for the use of gravity equations in policy analysis, and yields an imprecise point estimate of currency unions on trade close to one percent.

What determines trade patterns? Habit persistence in consumer tastes and learning-by-doing in production imply that history and culture matter. Deriving a dynamic gravity equation from a simple model, it is shown that cultural similarity is a product of history, so that trade patterns are a function of bilateral GDP, current trade costs, and the past history of trade costs. Using a trade data set which spans from 1870 to 2000, I demonstrate that many gravity variables operate via lagged trade, that historical trade shocks matter, and that trade patterns are persistent, even across centuries.