Hawks and Doves at War

• In Event: Domestic Politics and Foreign Policy

Abstract

Contemporary research on leaders and interstate conflict typically argues that leaders’ decisions are driven by either concerns about their political survival (Bueno de Mesquita et al. 1999) or their personal attributes (Horowitz, Stam, and Ellis 2015). There is something approaching a consensus that leaders’ desire to remain in power influences all aspects of conflict processes, from crisis initiation to war outcomes. In contrast, while there is increasing evidence leaders’ personal attributes influence the initiation and escalation of interstate crises, there has been comparatively little research on how they affect interstate war. I argue that leaders’ hawkishness, or their underlying willingness to use military force, influences not on only countries involvement in interstate crises, but also the prosecution and outcomes of interstate wars.

Leaders’ willingness to fight is central to our understanding of interstate war. However, existing research focuses on how the threat of ex post punishment influences leaders’ incentives to fight. Specifically, a domestic audience's willingness to sanction its leader is generally analyzed as a function of culpability or satisfaction with the war while its opportunity to punish a leader is typically analyzed in terms of regime type (Debs and Goemans 2010, Croco 2011, Croco and Weeks 2016). The empirical results yielded by these analyses motivate leaders' decision-making in countless theoretical explanations of interstate conflict.

I argue that leaders’ hawkishness should influence interstate wars through two pathways. The first is relatively direct. Hawks and doves systematically differ in their dispositional willingness to fight. This implies that, all else equal, hawkish leaders require more generous terms to reach a negotiated settlement than doves (Wolford 2007). The second pathway runs through domestic politics. Domestic audiences prefer hawks over doves when national security concerns are pressing, as is the case during crises and wars (Gadarian 2010, Mattes and Weeks 2019, Kiratli 2021). This results in domestic audiences disproportionately removing dovish leaders from power and leader transitions yielding relatively more hawkish leaders when structural conditions make interstate crises, and subsequently war participation, relatively likely.

The above argument implies leader hawkishness should influence important aspects of interstate wars. Domestic audiences’ preferences and differences in their underlying willingness to fight should result in hawkish leaders being more likely to be culpable for war than dovish leaders. Moreover, domestic audiences' preference for hawks during conflicts should result in relatively dovish leaders being disproportionately removed from office during hostilities. All else equal, this would result in the leaders in charge of countries at the end of wars being relatively more hawkish than other war-time leaders. Given the relationship between leaders’ willingness to fight has been shown to influence war termination and outcomes (Goemans 2000, Croco 2011), hawkish leaders should be less willing to reach a negotiated settlement and more likely to achieve a negative war outcome.

Initiation analyses using a recently developed measure of leaders’ latent hawkishness (Carter and Smith 2020) is consistent with my expectations during the period from 1919 and 2001. Specifically, I find that relatively hawkish leaders are less likely to be removed during a war and win a war, but are more likely to be culpable for war, in power at the end of a war, and lose a war. I plan on updating and expanding on my analyses in multiple ways. Perhaps most importantly, I plan on estimating how much of the relationships I’ve identified are driven by 1) the effect of leader hawkishness and 2) the non-random presence of relatively hawkish leaders in interstate wars.

Author

• Jeff Carter, Appalachian State University