Works In Progress

Democracy is in crisis, in retreat around the globe and under attack in International Relations (IR), as new scholarship challenges the “democratic triumphalism” that has dominated the field for the past two decades. We investigate democracy’s resilience a different way, re-analyzing all existing experiments on citizens and elites that randomize a country’s regime type (N = 205 estimates from 158 studies, fielded in 51 countries). We find that the attitudinal effects of democracy are remarkably consistent, systematically decreasing support for the use of force against a country, and increasing support for cooperation with it. These effects are homogeneous across respondent-level characteristics, have not appreciably declined over time, and weaken but do not disappear across space. Democracies are also perceived as more moral, less threatening, and more powerful, but critically, also less resolved. Regardless of whether satisfaction with democracy has declined at home, democracy appears to have maintained its allure when evaluating countries abroad.

Recent years have been marked by a surge in defections from cooperative arrangements, as states increasingly abrogate or violate international commitments. Yet, existing theories of international cooperation do not adequately explain the incentives and motivations of states as they formulate responses to such behavior. We argue that reputational dynamics are central to understanding how states respond to defection, and that reputation operates through two distinct channels… Read more

Scholarship on reputations in IR has left out a striking feature of human psychology: identity. Categorizing others as “us” or “them” is automatic, pervasive and has significant implications for reputations. We provide a framework—based on social identity theory—to explain how ingroup bias affects how reputations are generated and maintained… Read more

In the aftermath of foreign policy misadventures or domestic crises, it is common for observers to ask, “can [insert country]’s reputation be repaired?” As important a question as this is, scholarship on reputations in IR is not well-suited to answering it, as we demonstrate with a systematic review of past survey experiments on reputation in IR. We conceptualize this question as bearing on the elasticity, or sensitivity, of reputations to past behavior. Our broad argument is that reputations are subject to negativity bias—the tendency to privilege negative information over positive. More specifically, we argue that reputations are more easily damaged than repaired and that bad behavior is worse for states’ good reputation than good behavior is beneficial for bad ones. To demonstrate this, we first meta-reanalyze existing experiments (n=5, k=7) that are structured in such a way as to allow us to identify negativity bias in perceptions in IR. Second, we describe the results of an original pre-registered experiment that tests our argument.

One of the major methodological developments in experimental research in political science in the past decade has been the rise of cross-national survey experiments, in which researchers field experiments on samples from multiple countries. However, despite their newfound popularity, there is little sense of best practices, and the motivation for conducting cross-national experimental work is often unclear… Read more

While an older conventional wisdom held that reputation was “one of the few things worth fighting for,” later work suggested that reputations either did not matter or possibly did not exist at all. We argue that much of the recent “debate” over reputation has been hamstrung by scholars asking and answering entirely different questions about reputations. We introduce a new conceptual framework that divides reputational inferences into the evaluation phase (in which observers update their beliefs about type/attributes) and a diagnostic phase in which judgments are made about present and future behavior. We then introduce a new type of evidence into this debate, presenting results from original surveys fielded on four samples (including Israeli decision-makers, Israeli publics, and American IR scholars) to study the evaluation phase in which reputation costs were incurred by various actors as a result of the Russian invasion of Crimea and the Syrian civil war. Our results suggest that reputation costs exist, are worse in cases of failed threats, attach to both leaders and the countries they represent, and are perceived similarly by elite leaders and ordinary citizens. Finally, our studies revealed the existence of a “home-away gap”: compared to foreign observers, American IR scholars seem to underestimate the magnitude of reputation costs the US has incurred by backing down on threats.

Whether leaders taking the perspective of rivals or allies, student subjects taking the perspective of leaders in lab studies, or citizens taking their own perspective in hypothetical scenarios, most modern IR scholarship draws implicitly on perspective-taking. If our ability to engage in this sort of mental simulation were foolproof, we would have nothing to worry about. Unfortunately, several decades of psychological research suggests that individuals vary tremendously in their ability to do so, and that even when they can manage it, the effects are often negative. Ignoring this critical factor only detracts from our ability to generate and refine theories in IR or test our empirical predictions. We provide a conceptual framework for understanding perspective-taking in IR, focusing on the nature of the “target” (first or third-person) and individuals’ inability to adjust from their initial anchor: their own beliefs. Across three experimental studies, we find evidence that perspective-taking exacerbates pre-existing attitudes towards the use of force, making hawks more hawkish and doves more dovish. Perspective-taking thus makes people more like themselves, which raises the prospect that participants are less like themselves in studies that do not take perspective-taking into account.