Dyads Are Dead, Long Live Dyads! The Limits of Dyadic Designs in International Relations Research

Time and again, methodologists wrote papers with the potential to kill off dyadic designs in international relations research, only to pull back and—so long as one adopted the appropriate “tweak”—grant dyads a reprieve. Depending on the research question under consideration and the theoretical claims being evaluated, dyadic designs can provide valuable empirical insights. But these insights remain contingent on the researcher attempting to account for the limitations of dyadic data. To make this point, I concentrate on two difficulties for dyadic designs: the problem of interdependence and the problem of multilateral events. Drawing from empirical research on treaties and international organizations, I argue that a researcher need not abandon dyadic data so long as the researcher faces the former problem.