Of Ripple Effects and Reverberations: Disinformation in a Two-Level Game

Why, how, and to what effect do states use disinformation in their foreign policies? Inductive accounts variously address those questions, but International Relations has yet to offer a theoretical account. I propose Putnam’s two-level game (1988) as a candidate theory. A rationalist approach that jettisons the unitary actor assumption, the model accounts for previous accounts’ observations and suggests their interrelation and four overarching objectives. The model also generates novel implications about disinformation in foreign policy, two of which I test via separate survey experiments.The primary implication is that states can use disinformation to encourage polarization and in turn can reverberate into commitment problems. A survey experiment tests the first link in that chain, arguing that disinformation’s effects could be underestimated due to focus on belief outcomes; potential selection bias in active-exposure studies; and probable pre-treatment effects. It hypothesizes that passive exposure to novel political dis/misinformation has ripple effects on trust, affective polarization, and participation-linked emotions even among those that disbelieve it. It thus tests both the implication that disinformation can encourage polarization and that disinformation can be used to impact multiple potential outcomes at once. The second empirical paper tests the latter links in the disinformation-commitment problem chain. Building on a study that found U.S polarization decreases U.K. ally confidence (Myrick 2022), it argues that polarization uniquely increases chances of voluntary defection and does so not only due to government changeover risk but also weakened leader accountability. It employs a causal mediation analysis on survey experiment data to test whether a potential partner’s polarization increases their perceived unreliability and in turn decreases public cooperation preference. The commitment problem implication receives mixed support. The first experiment evidences no impact of partisan mis/disinformation on affective polarization, though that may be due to floor effects. The second experiment finds that polarization modestly increases perceived defection risk, but this increase is not necessarily strong enough to change public cooperation preference. Beyond those findings, the first experiment also uncovers that polarization may indeed have sociopolitical impacts on even those that disbelieve it, consistent with the multiple-outcomes implication.