This paper examines whether free trade agreements (FTAs) actually increase members' international trade. The gravity equation, a standard tool in international trade analysis, has been used to estimate the effects of FTAs on trade flows. However, the equation is subject to endogeneity bias, as trade policy is not exogenous. The authors use instrumental-variable (IV), control-function (CF), and panel-data techniques to address this endogeneity. While IV and CF approaches do not adjust for endogeneity well, a panel-data approach does. Accounting for endogeneity yields striking results: the effect of FTAs on trade flows is quintupled. The gravity equation typically includes variables such as GDP, distance, common language, common border, and FTA dummy variables. Previous studies have found mixed results, with some suggesting significant effects and others finding insignificant ones. The authors argue that the FTA dummy variable is endogenous, as countries may select into FTAs based on unobservable factors correlated with trade levels. This endogeneity bias leads to underestimation of FTA effects in traditional gravity equation estimates. The authors use panel data from 1960 to 2000 for 96 countries to estimate the effects of FTAs on trade flows. They find that traditional cross-section gravity equation estimates of FTA effects are biased and often underestimated. Using a theoretically-motivated gravity equation with differenced panel data, they find that the effect of FTAs on trade flows is significantly larger than previously estimated. The paper concludes that the gravity equation is subject to the same endogeneity bias as earlier studies, and that previous cross-section and panel estimates of FTA effects may be biased. The authors argue that the best method to estimate the effect of FTAs on trade flows is a theoretically-motivated gravity equation using differenced panel data, rather than instrumental variables applied to cross-section data. This approach provides more reliable estimates of the average treatment effect (ATE) of FTAs on trade flows.This paper examines whether free trade agreements (FTAs) actually increase members' international trade. The gravity equation, a standard tool in international trade analysis, has been used to estimate the effects of FTAs on trade flows. However, the equation is subject to endogeneity bias, as trade policy is not exogenous. The authors use instrumental-variable (IV), control-function (CF), and panel-data techniques to address this endogeneity. While IV and CF approaches do not adjust for endogeneity well, a panel-data approach does. Accounting for endogeneity yields striking results: the effect of FTAs on trade flows is quintupled. The gravity equation typically includes variables such as GDP, distance, common language, common border, and FTA dummy variables. Previous studies have found mixed results, with some suggesting significant effects and others finding insignificant ones. The authors argue that the FTA dummy variable is endogenous, as countries may select into FTAs based on unobservable factors correlated with trade levels. This endogeneity bias leads to underestimation of FTA effects in traditional gravity equation estimates. The authors use panel data from 1960 to 2000 for 96 countries to estimate the effects of FTAs on trade flows. They find that traditional cross-section gravity equation estimates of FTA effects are biased and often underestimated. Using a theoretically-motivated gravity equation with differenced panel data, they find that the effect of FTAs on trade flows is significantly larger than previously estimated. The paper concludes that the gravity equation is subject to the same endogeneity bias as earlier studies, and that previous cross-section and panel estimates of FTA effects may be biased. The authors argue that the best method to estimate the effect of FTAs on trade flows is a theoretically-motivated gravity equation using differenced panel data, rather than instrumental variables applied to cross-section data. This approach provides more reliable estimates of the average treatment effect (ATE) of FTAs on trade flows.