The importance of accounting for time trends when estimating the Euro effect on trade

To study the effect of the euro on international goods trade one typically estimates a panel model for the level of trade. Trade levels increase over time, and we show that this is not fully explained by the included regressors. Because the euro is only present at the end of the sample, this may have led to an upward bias in existing euro estimates to help explain the upward trend. To correct for that, we extend the panel model (a gravity model) by including a time trend that may have different effects across country-pairs. Data on industrialized countries over 1967-2002 show the existing euro effects of between 5% and 40% shrink to a statistically insignificant 3%. For comparison, the estimated trade effects of other currency unions are reduced from90% to 25%. Hence, accounting for time trends matters.