Do Some Men Want to Watch The World Burn?
When thinking about international security issues, the term “rational actor” is heavily misused. Proponents of hawkish measures against Iran suggest the Iranian government is fundamentally irrational—and thus would destroy itself to satisfy the dictates of its religiously influenced ideology. Doves counter that the Iranian regime has demonstrated a consistent survival instinct, and would hardly doom itself to destruction. Both sides fundamentally misunderstand half a century of research and modeling on rationality and its role in international security.
Rationality, unfortunately, is often construed as validity and soundness. If you are rational, you are intelligent, “civilized” (Western), organized, and risk-averse. The “irrational” is often visualized as Heath Ledger’s character, The Joker in The Dark Knight, a human random walk who scoffs at plans and simply wants to “watch the world burn.”
But rationality, as game theorist Phil Arena has consistently noted, is simply an assumption that an actor’s preferences are coherent enough to allow choices to be made. Let’s say the agent of interest “Ryan Evans” has a preference for spending a certain amount of time every week at a happy hour. Given that Ryan published a 4832-cell spreadsheet of happy hours, his preference for paying less for his alcohol isn’t likely to change in the near future. Because Ryan has the power to decide what choices will give him the best chance of achieving maximal happy hour attendance, he is rational.
However, Ryan is a Ph.D. student, works full time, and manages an online publication. Ryan, in other words, must not sleep a lot. Furthermore, Ryan must also base his decisions on the DC metro, an after-school educational program to help teach Beltway residents about the basics of quantum physics. So classical game theory assumptions about Ryan’s situational awareness, ability to calculate probability, and the level of computational cognitive processing power that our sleep-deprived student/think-tanker/online magazine publisher can bring to bear might not be appropriate.
So has Ryan suddenly become irrational—and do I need to assume that he will reject my WOTR pitches for completely arbitrary reasons? No. Bounded rationality is still rationality. Ryan still has his preferences and his ability to discern between them. And even if Ryan should decide to change his outlook on happy hours, there is a way to fit this into the overall scheme of games and decisions without assuming that Ryan’s preferences are completely random and incoherent.
Moving forward, there is an infinite number of ways we can incorporate rational assumptions about agents and relationships. And it isn’t just humans that we can look at. The invention of evolutionary game theory completely changed biology and ecology, which in turn is revolutionizing our understanding of how our own societies developed.
Finally, choice-based models also can be used to bolster the power of historical and case-based forms of analysis as well as detailed ethnographic work in the “human terrain. The choice-based tradition is vast, diverse, and flexibly suited to a large number of contexts, assumptions, and uses.
Choice-theoretic theories are far from perfect. The simplification choice-theoretic models require makes rigorous model validation paramount. Modelers also shouldn’t assume that mathematics will allow them to understand everything under the sun. You must calibrate the model to the specifics of the situation you are trying to understand. Game theorist and defense consultant Bruce Bueno de Mesquita will often extensively consult subject matter experts before deciding on actor preferences and game structure. And what psychological assumptions to make about the nature of choice and preference is a heavily controversial topic both among and within different communities of modelers.
So why is game theory relevant for security? Whether it’s the Soviet Union or modern-day Iran or al-Qaeda, decision-makers for years have been faced with analyzing opaque adversaries that they lacked good information about and found difficult to truly understand. Microfoundational theories that emphasized decisions in conditions of uncertainty had an immense appeal. While many simplifying assumptions had to be made, other assumptions requiring specific information and ethnographic thick descriptions that the U.S. would never possess were not necessary.
This is also why a true choice-theoretic understanding of rationality is so foreign to international security debates. It makes life difficult for political operatives seeking to construct broad claims based on an adversary’s “rational” status.
Want to go to war with Iran? Point to the regime’s grotesque mixture of religious fanaticism and totalitarian ideology as proof that it is the geopolitical equivalent of a mad dog that must be put down. Support a dovish approach to Iran? Argue that the regime’s behavior demonstrates an innate desire for regime survival.
The first flaw in these arguments lies in the assumption that the philosophical nature of an opponent’s goal is a reliable predictor of its near-term behavior. Iran’s leaders may, for all we know, hang a giant “What Would Ali ibn Abi Talib Do?” poster in the wall of their situation room. But any number of factors—such as the problems inherent in translating a religiously inspired worldview to the decidedly secular arts of strategy and statesmanship—complicate attempts to linearly extrapolate ideology to action.
And just because a geopolitical rival’s goal may include regime survival does not mean that we can necessarily rest easy. Incomplete information and uncertainty can often stimulate war. The nature of the game environment and number of players can turn brinksmanship into catastrophe. Finally, depending on adversary preferences, even a steep cost might be worth paying. Japan chose to gamble for the chance of surviving an American onslaught because the alternative was forfeiting hope of ruling the Greater East Asia Co-Prosperity Sphere. If we lose our empire either way, Japan reasoned, why not at least take the chance we could win?
Most importantly, choice-theoretic analysis of security is at its heart fundamentally relational. Actors in a game do not make decisions in a vacuum. Take the famous Prisoner’s Dilemma scenario, a game in which two cons must decide whether they will keep silent or rat out their partner. It’s a story about what we think someone else will do and what their hypothetical decision implies for our own choice. As Clausewitz notes in his metaphor of wars as “duel,” it is the strategic interaction that produces the overall outcome.
We can’t hope for a situation in which we might move away from simplified cultural or religious presumptions or crude and historically inaccurate theorizing about appeasement and resolve. The incentives are simply too great for actors in the American political game not to try to scare us with overly polemical notions about credibility, risk, and decision. But security analysts equipped with a better understanding about the nature of rationality will know what to do when it seems that some men just want to watch the world burn.
Adam Elkus is a PhD student in Computational Social Science at George Mason University and a columnist at War on the Rocks. He has published articles on defense, international security, and technology at CTOVision, The Atlantic, the West Point Combating Terrorism Center’s Sentinel, and Foreign Policy.