Abstract
The gravity model is widely used to estimate the effect of a reduction in trade costs on international trade flows. Researchers face many analytical decisions in their choice for an empirical strategy. This paper uses a specification curve analysis method to study the effect of these choices on estimated trade effects. I find that despite significant heterogeneity in estimated trade effects, the distribution in the empirical international trade literature overlaps fairly well with the estimates in this specification curve analysis. However, individual decisions have a significant impact on average estimated trade effects.