Power - The missing link in the trade conflict relationship

We join the debate on trade interdependence and conflict with a reexamination of Barbieri's (1996) intriguing empirical results showing that under certain circumstances, trade interdependence causes conflict. We argue that these findings resulted from a specification that was missing a variable. We return to Barbieri's models and introduce two independent power measures for countries within each dyad. When a correctly specified trade-conflict regression model incorporating the new power variables is performed, the constraining effect of interdependence becomes evident, and the results obtained are the reverse of Barbieri's.