At work here is the law of the instrument [that is, game theory and the Prisoner’s Dilemma]: give a small boy (or a researcher) a hammer and he will find things that need hammering.

As Kaplan (1964, 29) points out, often the problem is not that some techniques are pushed to the utmost, but that others may, in consequence, be ignored. Robert Axelrod, 1970¹

Because it captures the structure of a recurring sort of social predicament, there can be no doubt but that the prisoner’s dilemma model is bound to be of use to policy­makers...

Philip Pettit, 1985²

It is easy to appreciate how an American president, even the most conscientious, would be impaled on the horns of the intractable nuclear security dilemma. By all indications, the nuclear security dilemma should be resolved by assuring recipro­cal cooperation backed by potential devastating retaliation, in keeping with the classical liberal adage of war for war and peace for peace. However, rational strategists considered the threat to harm millions of innocent people once deter­rence had already failed to be both immoral and irrational. In 1978, Gregory Kavka, who contributed the most rigorous philosophical analysis of this problem at that time, had little to offer President Jimmy Carter to place the strategy of

¹ Robert Axelrod, Conflict of Interest: A Theory of Divergent Goals with Applications to Politics (Chicago: Markham, 1970), 6; Abraham Kaplan, The Conduct of Inquiry (Chandler, 1964).

² Philip Pettit, “The Prisoner’s Dilemma and Social Theory: An Overview of Some Issues,” Australian Journal of Political Science (1985) 20:1, 1-11, at 1.

143

mutual assured destruction on firm ground, save that he should leave the role of issuing credible deterrent threats to agents who were either less than moral or less than rational with the hope they would never be carried out.³ In 1980, Kavka renewed his efforts to defend the minimum deterrence of MAD against the countervailing war-fighting posture of NUTS. He recommended that the United States should disambiguate its intentions from those characterized in the Prisoner’s Dilemma model of the nuclear security dilemma and arms race by clearly demonstrating the intention to cooperatively avoid nuclear war. He recommended building trust by offering assurance of intentions to follow through on cooperation if assured reciprocity rather than demonstrating the preference to secure dominance consistent with PD logic.

Leaving aside for the moment how best to resolve the paradox of nuclear deterrence, in the 1980s philosophers and social scientists came to accept that much more mundane contexts - most prominently the social contract, collective action, and market exchange - were best modeled with the Prisoner’s Dilemma. This model no longer was reserved for worst-case scenario planning; instead, the fact that the Prisoner’s Dilemma had proven to be useful in analyzing nuclear deterrence and the potentially apocalyptic arms race was used as an advertisement for its relevance.⁴ One of the greatest proponents of the game theoretic approach to studying social interactions, Robert Axelrod, moved swiftly from the Cold War security quandary to individual behavior among friends and business partners and the riddle of how to exit the state of nature associated with Thomas Hobbes’s anarchy. He begins his highly acclaimed and influential book The Evolution of Cooperation with the following questions:

When should a person cooperate, and when should a person be selfish, in an ongoing interaction with another person? Should a friend keep providing favors to another friend who never reciprocates? Should a business provide prompt service to another business that is about to be bankrupt?⁵

Axelrod prepares to offer advice to anyone confronting these questions and bases the credibility of his wisdom on its relevance to nuclear security.

³ This is what Kavka concludes in “Some Paradoxes of Deterrence,” Journal of Moral Philosophy (1978) 75:6, 285-302 and what Douglas Lackey reports in that other moral philosophers had gravitated toward accepting the minimum deterrence characteristic of MAD, yet remained stymied by the fact that only those manifestly willing to use nuclear weapons in line with NUTS could maintain credible deterrence by the 1970s, “The American Debate on Nuclear Weapons Policy: A Review of the Literature, 1945-1985,” Analyse and Kritik (1987) 9, 40-43. This Lackey literature review is essential reading in the nuclear security debate and makes clear the portal both international relations theory and political theory went through conjointly in tackling the problem of nuclear security with the preeminent tool of rational deterrence theory, which is equivalent to game theory.

⁴ This is evident in one of the most important texts discussing the PD, Richmond Campbell and Lanning Sowden, Paradoxes of Rationality and Cooperation (Vancouver: University of British Columbia Press, 1985).

⁵ Robert Axelrod, Evolution of Cooperation (New York: Basic Books, 1984), vii. to best elicit cooperative behavior from the Soviet Union?”^[351] Building on his 1970 Conflict of Interest, Axelrod’s application of strategic rationality spans the Cold War security dilemma, the broader sphere of international relations, market exchange, governance and the social contract, interpersonal relations, and even evolutionary biology.

Axelrod, along with frequent coauthor and neoliberal institutionalist Robert Keohane, was a key figure in establishing the neoliberal approach to political economy. Neoliberalism provides an explanation for how narrowly self-interested rational actors may achieve cooperative outcomes under the assumption of strategic rationality.

Axelrod links his research to studies of cooperation and altruism in evolu­tionary biology and made connections between traditional Western political theory and game theoretic modeling. He is quick to draw attention to the apparent confluence of three centuries of thought under the single rubric of the Prisoner’s Dilemma: “Hobbes regarded the state of nature as equivalent to what we now call a two-person Prisoner’s Dilemma, and he built his justifica­tion for the state upon the purported impossibility of the sustained cooperation in such a situation.”^[355] Throughout his work, Axelrod articulates the intellectual foundations of neoliberal markets and governance in the Prisoner’s Dilemma social contract.^[356]

The reinvention of the social contract in keeping with strategic rationality and the Prisoner’s Dilemma model of the security dilemma have been coex­tensive with the emerging practice of neoliberal economics, and viewing behavior to be programmed into agents, therefore modeling presocial organ­isms in evolutionary contexts on par with human subjects (Part III).

Before moving into the chapters in this Part II, readers are invited to first consider a central question: how did the worst-case planning that was deemed suitable to, if not the factual realities, then the logical intricacies of rational deterrence become transferred into the interiority of civil society to produce nuclearized sovereignty? How did strategic rationality, which typically assumes consequentialism (only outcomes matter), realism (value exists prior to social relations), and hyper-individualism (“other-regarding” signifies viewing others as strategic maximizers like oneself), come to be the only approach to coherent action available to individuals throughout their lives?^[361] The pivotal point of paradigmatic shift occurred when the analytic philosopher and political theorist David Gauthier extended the anxiety over the incredibility of immoral threats underlying mutual assured destruction to question the coherence of moral promises and commitments.

Gauthier’s 1984 essay “Deterrence, Maximization, and Rationality” marked the point at which analytic political theory fully embraced rational decision theory and endorsed the PD view of governance, thereby fully breaking with the classical liberal world of reciprocal respect for one another’s right to exist, enjoyment of property as an extension of personhood, and autonomy to enter into private agreements without the need for invasive government mandating compliance. Gauthier’s work accepts that rational individuals must comport with the axioms of decision theory; no other logics of action are available to them. The neoliberal subject can expect to be caught in Prisoner’s Dilemmas at any moment and will need to devise means to resolve them using the tools of noncooperative game theory. The classic liberal world of reciprocal no-harm and voluntary compliance with agreements made recedes because strategic rationality does not condone behavior that does not augment each individual’s expected utility in consideration of achieved outcomes. This neoliberal alienation of selfhood is even more complete than in the classic liberal’s proposal that freedom is equivalent to selling one’s labor on the open market.1⁷ No domain of relations or experiences can transcend the single criterion imperative to trade off every eventuality against another and make choices accordingly throughout the supergame of each individual’s life. Individuals are free to do whatever they please, subject to the constraint of their willingness and ability to pay for their choice. Instead of market relations being privileged practices contained within the rule of law, building on security to achieve positive-sum growth and experiential cultural value that may exceed limited physical resources by nonlinear measures, expected utility theory and game theory are the rules that all life forms must successfully navigate to survive and propagate.¹⁸ Individuals are free to act on any choice contained in their set of opportunities. The volitional quality of choice is eroded in favor of strategic rationality, which was originally designed as a logic of action that could provide a complete set of instructions needing no conscious implementation.¹⁹ Stress is perpetual because one’s status is never secure as a result of uncer­tainty, risk, and other actors’ combative stances and likely adoption of coer­cive bargaining tactics.

Gauthier’s work represents a milestone because he definitively explains how the puzzle of achieving credible deterrence is equivalent to satisfactorily resolving a “whole range of situations, including most prominently general­ized Prisoners’ Dilemmas.”²⁰ He concludes that the PD mirrors not only the nuclear security dilemma but also both the challenge of maintaining social order in a civil society and the problem of conducting amicable exchange: “The state of nations and, more especially, of nuclear powers is our nearest analogue to the state of nature, and Hobbes’s advice applies to it.”^{[362] [363]}1 In this single densely argued article, Gauthier takes Thomas Schelling’s twin admit­tance of the complete validity of strategic rationality and its central analytic problematic Prisoner’s Dilemma from the domain of worst-case preparedness into the heart of the social contract. Thus, relationships among individuals and between individuals and the state all fall under the umbrella of strategic rationality. Its appellation “rational choice” softens its strategic and comba­tive character, instead emphasizing individual’s choice between being rational or irrational.

In his work, Gauthier attempts to solve both the paradox of nuclear deterrence and the problem of Hobbesian anarchy, which he claims spans international relations and civil society, by finding a way out of the Prisoner’s Dilemma. He credits Schelling for first identifying that the PD best models the nuclear security dilemma. However, rather than following Schelling’s and Kavka’s recommenda­tions to exit the PD into the more favorable Stag Hunt game, Gauthier turns his attention on cutting through the Gordian knot of the PD’s inevitable suboptimal, mutually impoverishing outcome. Gauthier’s research is at the epicenter of the rupture between classical liberalism and neoliberalism because he fully embraces both strategic rationality and its entailed paradox that self-interest leads to mutual disaffection in the Prisoner’s Dilemma.

Gauthier’s analysis identifies two means of exiting the PD. On the one hand, if one can issue a credible threat to retaliate should the other fail to cooperate, then the other will cooperate, and the PD is solved. The other way out of the PD, so Gauthier reasons, is to credibly promise that one will cooperate if the other does as well. Neither solution is complete without both parties adopting the proposed solution symmetrically: each would need to deter the other, and each would need to credibly promise to cooperate if the other does. Deterrence and assurance have the same puzzling structure because in both cases indivi­duals offer to take an action that at the time of its execution fails to have any causal import in realizing their preferences. From the perspective of strategic rationality, because the act serves no instrumental function in promoting the actor’s goals at the time it is performed, the act must be irrational. Therefore, neither should the rational actor implement the act, nor is it credible that either the deterrent threat or assuring promise has any weight in averting the mutual defection outcome of the Prisoner’s Dilemma.

Gauthier’s analysis follows from equating solving the PD game by assuring the other of one’s intent to cooperate in solving it by compelling cooperation via the threat of retribution. Thus, he equates a credible promise to cooperate with a credible threat to retaliate. In both cases, the PD structure of the problem accepts that each most prefers to sucker the other. The Prisoner’s Dilemma reinforces the credibility of threatening the opposition in the case they fail to cooperate because, as with nuclear deterrence, the United States adopts prefer­ences consistent with defecting and punishing the opposition regardless of the Soviets’ cooperation or defection. However, following through on the punishment in the case of nuclear deterrence signifies both murder and suicide. This parallel construction of promising and threatening underscores the ques­tion of how, in either case, the commitment to follow through on a statement of intent can be credible if it lacks the casual power to realize the agent’s expected gain at the time the action is taken. Specifically, if I threaten to retaliate if you fail to cooperate, but the act of retaliation does not further my goals or max­imize my gain, then after you fail to cooperate, punishing you serves no purpose for me.²² Similarly, so Gauthier argues, if I make a promise to exchange money for a good, once I have the good, and delivering the money serves no causal purpose in maximizing my expected gain, then I have no instrumental reason to fulfill my commitment. In both cases, the utility maximizing agent will not follow through on either the threat or the promise.

Both Kavka and Gauthier agree that the United States must be perceived as first and foremost a moral nation that would eschew acting on the initiative of gross brutality, specifically in the case of the failure of deterrence. However, their analyses hone in on different aspects of the paradox of deterrence. Kavka persistently emphasizes that moral scruples prohibit actors from issuing cred­ible deterrent threats. Gauthier stresses that the causal inefficacy of the punitive retaliation means that perpetrating punishment fails to maximize the expected utility of the perpetrator and thus is irrational to enact and impossible to credibly threaten.

Whereas Kavka suggests that the assurance solution of the Prisoner’s Dilemma may be possible because of the integrity of choice consistent with the moral institution of promising and the underlying hope to achieve mutual cooperation despite actors’ presumed temptation to sucker the other, Gauthier concludes that both immoral threats and moral promises are contrary to stra­tegic rationality. In game theory, expected utility only pertains to outcomes, which typically track an instrumentally salient property of the world evident to intersubjective inspection. Thus, if expected utility in games only registers the value of outcomes, and every consideration of worth must be reflected in the agents’ expected utility function, then exiting the Prisoner’s Dilemma by pro­mising the other agent of one’s intention to cooperate if assured reciprocity must fail. After the other actor cooperates, the PD preference matrix, which contains all the information relevant to individual choice, still requires the rational actor to defect.

Thus, Gauthier and the philosophers working in the rational choice tradition concurred that contracts from dyadic exchange through to complex multi-actor social contracts are best modeled as Prisoner’s Dilemmas and resist solution by rational actors. Agents who comport with game theoretic instrumental ration­ality are caught in an undertow because they “cannot reason themselves into being rule followers [who maintain commitments entered into either by agree­ment or tacit consent] and yet rule following seems absolutely critical to social life.”²³ The adoption of strategic rationality, as a result, contradicts social practices heretofore deemed sensible and fundamental to social concourse. Thus, in confronting the nuclear security dilemma wherein initially deterrent

²² Footnote 13 applies equally here.

²³ Gerald Gaus, The Order of Public Reason (Cambridge: Cambridge University Press, 2011), 146-147.

threats that compromised the moral character of the United States were deemed wrong for intending harm, hence incredible and irrational, subsequently moral assurances were ensnared because they contradicted the structure of strategic action.

The nuclear security dilemma opened the door to a world in which the rational autonomous decision maker obeys the axioms of expected utility and rules of strategic choice just as the nuclearized superpower exercises sovereignty through compelling and deterring other nations. Agents accept that the same modus operandi of incentives and outcomes, as opposed to also considering legitimacy and procedures, delimit all intelligible relations. Part II investigates the implications of viewing the rationale for government in terms of strategic rationality and Prisoner’s Dilemma. Chapter 5, “Hobbesian Anarchy,” dis­cusses how the rational choice reading of Hobbes’s Leviathan and its PD model of anarchy leads theorists to consider that coercive force is the necessary and sufficient condition for achieving civil society instead of as a necessary condition to be complemented by consent and self-restraint. Chapter 6, “Social Contract,” follows James M. Buchanan’s and John Rawls’s debate over fair play and the motivational weight of hypothetical consent to terms of distribution. Buchanan’s neoliberal account of government privileges status quo power relations and cannot exit from the realm of coercive bargaining. Arguing that all people have the same temptation to transgress law independent of their means or inclusion in the social contract, Buchanan suggests a justifica­tion for a maximal security state consistent with mass incarceration. Chapter 7, “Unanimity,” argues that Buchanan’s attempt to derive constitutional legitimacy from unanimous agreement is undermined by the lack of motivational content of strategically rational consent. Chapter 8, “Consent,” follows Richard Posner’s further demolition of consent by his equation of the legal weight of ex post and ex ante consent. Posner’s theoretical position has the implication of permitting that better-off individuals can acquire resources from less well-off individuals by paying compensation after harm without prior consent. Lastly, Chapter 9, “Collective Action,” analyzes how the Prisoner’s Dilemma model of joint action fails to distinguish between the challenges of the temptation to free ride and resignation over the causal negligibility of contributing, in situations classified as large-scale collective actions or “tragedies of the commons.” The conflation of the hope to cheat with the concern that one’s actions cannot make an appreciable difference makes it difficult for either political theorists or institutional designers to establish effective regulatory regimes.

Together, these chapters articulate the implications of a consistent rational choice approach to political theory that yields a basis for neoliberal political economy in stark contrast to the classical liberal state of mutual respect and minimal intervention. Viewing rational agency in terms of expected utility and strategic rationality creates a tendency to measure all value in terms of a zero­sum finite resource and to accept as nonnegotiable the imperative to maximize gain individualistically as the primary means of survival. Free riding is not one potential outcome, but the rationally condoned result.^[364] Coercive bargaining too is consistent with rational choice and cannot be surmounted. The strength of neoliberal political theory is that it embraces value pluralism and rejects paternalism.^[365] However, accepting that individuals might be narrowly self­interested, unprincipled, and uncooperative is different from teaching that rationality implies this behavior.