World trade has grown exponentially during the last 60 years. Admittedly, it is not clear if this development can be assigned to international trade agreements like the World Trade Organization or the Generalized System of Preferences as previous empirical studies found contradicting results. In this paper we generalize the different approaches used in the lit- erature to estimate the role of GATT/WTO and the Generalized System of Preferences for trade. We use a gravity model and apply FE estimation on a disaggregated bilateral data set of the trade flows between 145 countries across 1962-99. In our analysis we find a significant positive effect of WTO membership on bilateral trade. Referring to other multi- lateral institutions, we find robust evidence that membership in regional trade agreements or currency unions substantially increase bilateral trade flows as well. By contrast, we fi nd that the Generalized System of Preferences does not foster trade in general, rather the opposite. This might be due to the opportunistic behavior of industrial countries that grant GSP schemes as long as the concerned products are relatively unimportant, but restrict them as soon as they become relevant.
WTO GSP regional trade agreements currency union gravity model international trade