“…To correct this bias, several methods may be used: (a) the instrumental variable method with a two‐stage estimation (e.g. Chen & Ravallion, ; Magee, ), but this method requires finding plausible and significant instrumental variables; (b) the matching method (Baier & Bergstrand, ; Smeets & Warzynski, ); (c) the difference‐in‐differences matching (Egger, Egger, & Greenaway, ; Girma and Görg, ); (d) Heckman control function method or Helpman, Melitz, and Rubinstein () method (here, the problem is to determine the “exclusive variable”); and (e) the fixed‐effects method (Coulibaly, ; Gil‐Pareja et al, ). Many researchers implement the gravity equation with EIA dummy variables and fixed effects to find consistent and precise empirical estimates of the average treatment effect of EIAs on trade flows (Anderson & Yotov, ; Head & Mayer, ; Eicher, Henn, & Papageorgiou, ).…”