We Don't Do That: National Identity and Nuclear Non-Use

In international relations scholarship, political scientists have often taken after microeconomic theory in assuming that states are rational and selfish (Waltz 2001). Neorealist scholars presuppose that states are strictly security seeking, unmoved by considerations of fairness or altruism in an international system predicated on anarchy and self-help. Since the “constructivist turn,” however, scholars have challenged the descriptive accuracy of assuming states to be egoistic and the international system anarchic (Wendt 1999; Finnemore and Sikkink 1998; Wendt 1992). Instead, individual leaders, domestic constituencies, and states as units in and of themselves can be constrained by norms of how they are meant to behave. Crucial to how norms are theorized to affect state behavior is identity. If an action is so morally reprehensible as to contradict a state’s perception of itself, then it will refrain from whatever that action is. Because identity and norms are difficult to measure numerically, constructivist scholars have often resorted to process-tracing and other qualitative means to substantiate their argument that states’ behavior is mediated by ideational factors (Peez 2022). Although these approaches have been useful in illustrating how identity affects elites’ attitudes toward nuclear use (Tannenwald 2007), collateral damage (Crawford 2013), and other norm-breaking behavior, it has limited the scope of constructivist epistemology. The lack of quantitatively-driven, positivist study is striking given that experimental research has shown priming national identity can alter political behavior, such as by reducing affective polarization (Levendusky 2018) or changing attitudes toward immigration (Wojcieszak and Garrett 2018). In this paper, I seek to redress this methodological gap by fielding a survey experiment in which Indian respondents are exposed to identity-based arguments opposing a hypothetical nuclear attack against Pakistan. My focus on nuclear weapons follows from scholarship that has characterized non-use as among the strongest norms in the international system. Describing use as repulsive and “taboo,” Nina Tannenwald (2007) contends that non-use is reflexive. Even in dire circumstances, a shibboleth is attached to nuclear weapons, effectively eliminating states’ motivation to use them. Realists have criticized Tannenwald’s argument, saying that if a norm against non-use exists at all, it is epiphenomenal—the product of egoistic states acting out of self-interest (Sagan 2004). To substantiate their argument that no taboo exists, Scott Sagan and Benjamin Valentino have conducted survey experiments revealing that Western public opinion is far more receptive to the use of nuclear weapons than previously thought, with compelling majorities approving of their use in circumstances where they achieve a compelling strategic objective (Press, Sagan and Valentino 2013; Sagan and Valentino 2017; Dill, Sagan and Valentino 2022). By focusing on India, I test the salience of identity-based arguments in the circumstance where they are most likely to have an effect. Unlike in Western, nuclear armed states, political elites in India have historically framed non-violence as morally obligatory and a touchstone for its foreign policy (Perkovich 2001; Hymans 2006; Mehta 2009; Bajpai 2014). Following World War II, India was among the staunchest proponents of global disarmament and continued to be so even after it tested nuclear weapons. Today, although there is widespread doubt that India would abide by its NFU during a crisis (Clary and Narang 2019; Sundaram and Ramana 2018), political elites still explicitly assert that first use is antithetical to Indian culture and civilization. Aside from entertaining how identity affects attitudes toward nuclear weapons use, I consider how the institutionalization of non-use can do the same. Whereas previous studies have focused on how international law can restrict support for nuclear use (Carpenter and Montgomery 2020; Sagan and Valentino 2024), I consider how no first use policies (NFUs) do the same. Whereas international law promotes shared norms, NFUs reinforce perceptions of states’ personal identities as being incompatible with nuclear use. This line of argument presents a novel modification to scholarly understanding of NFUs’ credibility. Specifically, it implies that NFUs can trigger audience costs by reifying an image of state identity as being incompatible with nuclear use. Ultimately, the purpose of this thesis is twofold. Narrowly, it contributes to the upswell of survey experiments from the past ten years, in which scholars have sought to understand what factors impact public support for the use of nuclear weapons. It does so by interrogating the effects that national identity and institutionalization can have, and whether those effects are significant in the manner that scholars of the taboo suggest. More broadly, it seeks to evaluate claims in constructivist literature that norms can have a “constitutive effect,” in which people’s foreign policy preferences are impacted by their impression of state identity.