HOBBES’S LEVIATHAN AND THE PRISONER’S DILEMMA

Many theorists tackling the rationale for government through the worldview of rational choice propose that Hobbes’s state of nature is best represented by a Prisoner’s Dilemma game.1⁴ There are many possible reasons for such an argu­ment.

Regardless of which motive propels actors - fear, gain, or glory - game theory seems to capture the problem of cooperation. Game theorists collapse these three motives, which Hobbes distinguishes, into a single criterion payoff matrix.¹⁶ Additionally, game theorists must assume that actors are aware of all conceivable outcomes, that they have consistent preferences over these potential outcomes across their lifetimes, and that they can accurately estimate relative probabilities of outcomes coming to pass. Once one concedes these assump­tions, and that the scenarios that actors repeatedly face reflect the paradigmatic Prisoner’s Dilemma, then Hobbes’s argument for Leviathan seems appropriate. The Prisoner’s Dilemma best represents the state of nature in which individuals confront one another. The state must step in to alter actors’ incentives so that they can achieve the mutually desirable outcome of cooperation versus mutual attack. According to this reading of Hobbes, the state is particularly required because even if actors agree to cooperate, each still has the most favored temptation to defect.

And covenants, without the sword, are but words, and of no strength to secure a man at all. Therefore notwithstanding the laws of nature [self-preservation, and the predilection to seek peace once self-preservation is assured]... if there be no power erected, or not great enough for our security; every man will, and may lawfully rely on his own strength and art, for caution against all other men.¹⁷

Game theorists concur. Agreements without binding force behind them are mere words or in the parlance of game theory, cheap talk. The primary problem of social order then is that given the myriad situations throughout civil society in which agents’ preferences and choices are surmised to resemble that of a Prisoner’s Dilemma, everyone prefers mutual cooperation to mutual defection, yet everyone prefers sole defection to unanimous cooperation.¹⁸ An enforce­ment body must therefore be introduced to alter the otherwise tempting payoff for attempting unilateral defection.

This straightforward reading of Hobbes may be referred to as Rational Choice Hobbes. Many theorists have developed this interpretation of Hobbes’s Leviathan and the challenge of achieving social order.1⁹ It now stands as a canonical reading of Hobbes as well as a standard game theoretic argument explaining the need for government. In their authoritative game theoretic text, Hargreaves Heap and Varoufakis move directly from introducing the Prisoner’s Dilemma game to using it to provide the rationale for government. The difficulty is that in every situation characterized by the Prisoner’s Dilemma preference matrix, all would agree to mutual cooperation but then would look to exclude themselves from carrying through as promised.²⁰ Hargreaves Heap and Varoufakis conclude that “there needs to be a mechan­ism for enforcing agreement. Hobbes’s ‘sword’, if you like.” Crucially, they add, “It is this recognition which lies at the heart of a traditional liberal argument, dating back to Hobbes, for the creation of the State (or some enforcement agency to which each submits).”²¹

The same problem of how to achieve compliance with an agreement made in a state of nature resurfaces within a nation-state.²² Individuals must pay taxes to enable the state to function.

¹⁹ In addition to those mentioned in note 14, see James M. Buchanan, Limits of Liberty: Between Anarchy and Leviathan (Liberty Fund, 2000); Hargreaves Heap andVaroufakis, Game Theory, 2004, 34-38.

²⁰ David Lewis addresses this characteristic predicament in his discussion of conventions vs. social contracts: in a social contract, individuals may prefer for others to cooperate while they themselves engage in lone disobedience, whereas under conventions all individuals prefer mutual cooperation to sole defection as in an Assurance Game; Convention, 1969,90-91.Lewis offers a game theoretic interpretation of Hobbes’s resolution of the social contract, suggesting that (consistent with preferring peace to war), actors’ preferences will reflect mutual cooperation over unilateral defection, and that therefore, “these accepted obligations [in the form of “tacit consent or fair play” to honor the sovereign’s commandments] will be counted as a component of preferences, not as an independent choice-determining force”; Convention, 2002, 94. However, Lewis’s resolution of Hobbes’s state of nature into the social contract misses that in orthodox noncooperative game theory using expected utility theory over outcomes not only gives competition over scarce resources the payoff structure of a Prisoner’s Dilemma, but more­over leaves no latitude “to think of someone’s preferences as the resultant of all the more or less enduring forces that go into determining his choices,” which would permit “the modification of our preferences by obligations” that exist because individuals prefer peace over war, 93 and 94. Thus, Hobbes’s three analytic laws of nature prescribe obligation, which then transforms preferences contradicting orthodox noncooperative game theory. For a clarifying discussion, see Amartya K. Sen, who observes that “the rejection of self-goal choice reflects a type of commitment that is not able to be captured by the broadening of the goals to be pursued...

²¹ Hargreaves Heap andVaroufakis, Game Theory, 2004, 174; note that Phillip Pettit concurs with this argument for law enforcement, “The Prisoner’s Dilemma and Social Theory,” 1985.

²² Taylor, Possibility of Cooperation, 1987, 125-179, clearly articulates this, and he argues that conditions under the state will only exacerbate the PD structure of cooperating, see 167. threat of sanctions. Thus, running the state is likewise conceived as a multi­person Prisoner’s Dilemma game.²³ Gregory Kavka articulates the PD propor­tions of the problem. Security is a public good, which is defined as a resource supplied as a result of private contributions that everyone has access to regard­less of whether that individual actually contributed. Kavka explains, “Most any individual faces a situation similar to multiparty prisoner’s dilemma... Because contributing is not necessary for receiving the benefit, relatively few will con­tribute relatively little, and the good in question will be undersupplied.” According to this PD analysis, although everyone seeks peace, everyone prefers to forgo contributing while others pay. The solution is taxation at the point of the sword. Kavka concludes, “financing of the good by enforced taxation can eliminate the prisoner’s dilemma aspects of the situation and produce the socially desirable outcome. Hobbes realized this as regarded the public goods of national defense and safety from attack by one’s fellows.”²⁴ In his analysis of Hobbes’s argument for absolute sovereignty, Michael Moehler agrees that the “problem of compliance is best modeled by a one-shot PD game.”²⁵ Economists also join in on this understanding of the rationale for government.²⁶ It is difficult to overemphasize the extent to which contemporary theorists informed by rational choice theory accept that the social contract and government are challenges best represented by the Prisoner’s Dilemma model.²⁷

The rational choice understanding of Hobbes is easy to present.

The game theoretic analysis of the problem underlying the achievement of social order identifies a Prisoner’s Dilemma at three points: in the state of nature

²³ Taylor explains this, ibid., 127-148; see also his Anarchy and Cooperation, 1976.

²⁴ Kavka, Hobbesian Moral and Political Theory, 1986, at 246 (and preceding quote).

²⁵ Michael Moehler, “Why Hobbes’ State of Nature Is Best Modeled by an Assurance Game,” Utilitas (2009), 21:3, 297-326, 311.

²⁶ See Dennis Mueller, Public Choice III (Cambridge: Cambridge University Press, 2003), 9-12.

²⁷ Brian Skyrms is an important exception; see The Stag Hunt, 2004.

²⁸ Taylor observes that “individual preferences in Hobbes’s state of nature have the structure of a Prisoner’s Dilemma game at any point in time” [his emphasis], and that actors are faced by two choices, to cooperate or defect, The Possibility of Cooperation, 1987, at 134.

figure 10.

figure ιι. NeoliberalGovernance

prior to the establishment of the sovereign state, motivating the rationalization for the state, and within the state because actors will seek to defect from both private agreements and contributing to the tax base of the state. The enforcement power of the state is fundamental to establishing and maintaining the social order game theorists envision, and they pinpoint Hobbes as a key pioneer reaching this conclusion. This opinion has prevailed despite the fact it both contradicts Hobbes’s classification as an early liberal political theorist and places neoliberal political theory on a bleakly authoritarian footing connoted by this heavy-handed solution to the problem of achieving social order.^[375] Game theorists are aware that somehow the rational choice diagnosis of the challenge of maintaining governance is a more onerous task than Hobbes foresaw. One textbook acknowledges that for Hobbes the state would be absolute, but “its interven­tions... quite minimal.” In contrast, the game theoretic PD analysis “seems to suggest that the State... will be called upon to police a considerable number of social interactions... the boundaries of the State... will be drawn quite widely.”^[376]