Date of Award

Fall 12-2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


International Studies

Committee Director

Jesse Richman

Committee Member

Regina Karp

Committee Member

John Sokolowski


How does structure shape behavior and outcomes in crisis bargaining? Formal bargaining models of war rely on expected utility theory to describe first-order effects, whereby the payoffs of war determine actors’ “resolve” to fight as a function of costs and benefits. Value preferences of risk and future discounting are routinely treated as predefined and subjective individual attributes, outside the strategic context of bargaining or independent from expected utility. However, such treatment fails to account for context-conditional preferences sourcing from actors’ expectations of relative gain or loss. Drawing on a wealth of experimental evidence from behavioral economics, but without departing from rational choice or compromising theoretical parsimony, this dissertation proposes a systematic differentiation of value preferences conditional on anticipated gain/loss, i.e., the endogenous shift in power bargaining is expected to produce. Whereas the utility of gain incentivizes a challenge to the status quo, the disutility of loss imposes reactive resolve via asymmetrical risk-acceptance and lower discounting of future payoffs. The proposed theory of reactive resolve, thus, reveals the second-order impact of structural conditions on behavior and outcomes in crisis bargaining. Short of this behavioral effect, bargaining models exhibit a tendency of automatic adjustment of benefits which fails to capture the very essence of conflict and encourages erroneous hypotheses about the role of superiority, such as the nuclear superiority hypothesis reviewed and rejected as part of this research.

The prescriptive and predictive inaccuracy of the standard rationalist approach is evident in the solution of the most fundamental bargaining problem - a credible commitment problem arising in the context of “bargaining over future bargaining power” (Fearon 1996). By formally integrating and simulating expected gain- and loss-induced preferences, this study demonstrates substantial deviations from previous results. Based on the findings, several theoretical and empirical implications are derived concerning the mechanism of crisis escalation, the relationship between the distribution of power and the likelihood of war, and the challenge to coercion. The prescribed mechanism is then empirically tested against cases of compellence and deterrence, including two of the most significant cases of nuclear crisis, using process tracing as a qualitative tool of causal inference.