II Tit for Tat
[Robert] Axelrod, like many political scientists, economists, mathematicians and psychologists, was fascinated by a simple gambling game called Prisoner’s Dilemma. It is so simple that I have known clever men misunderstand it completely, thinking there must be something more to it! But its simplicity is deceptive.
Whole shelves in libraries are devoted to the ramifications of this beguiling game. Many influential people think it holds the key to strategic defense planning, and that we should study it to prevent a third world war. As a biologist, I agree with Axelrod and [W. D.] Hamilton that many wild animals and plants are engaged in ceaseless games of Prisoner’s Dilemma, played out in evolutionary time.Richard Dawkins, 20091
Since the Prisoner’s Dilemma is so common in everything from personal relations to international relations, it would be useful to know how best to act when in this type of setting.
Robert Axelrod, 19842
Noncooperative game theory has been applied to nuclear strategy, the social contract, public goods, and also evolutionary biology. Everywhere its logic is the same: optimization, or expected utility maximization, occurs on the level of individuals in a population who typically compete for scarce resources. The Prisoner’s Dilemma game is emblematic of the perceived problem of cooperation: individuals seek propitious outcomes, but ultimately prefer their own gain, even at the expense of someone else’s loss. Having generated the concept of an evolutionary stable strategy (ESS), evolutionary game theorists deduce that every population of actors must be impervious to a deviant member designed to exploit others (see Chapter 10). To protect against individuals’ exploitation by other actors, natural selection eliminates acts of gratuitous altruism because they would undermine its perpetrators’ survival chances, thereby eventually
1 Richard Dawkins, The Selfish Gene (Oxford: Oxford University Press, [1989^2009), 203.
2 Robert Axelrod, The Evolution of Cooperation (New York: Basic Books, 1984), 17.
269 eliminating individuals disposed to this type of behavior. Even if group selection occurs among human societies, still every highly cooperative population must be impervious to individualistic exploiters. This, it is hypothesized, requires individualistic selfishness.
The Prisoner’s Dilemma game has proven to be endlessly fascinating for representing the foil against which cooperative behavior must test its viability. Given evolutionary biologists’ interest in the material conditions of survival, that is, caloric intake and wherewithal to procreate, the tendency of organisms to waste resources in the suboptimal Nash equilibrium of the Prisoner’s Dilemma inspired copious research. It seems perplexing that animals competing for survival would repeatedly defect in Prisoner’s Dilemmas, or in Stag Hunt and Chicken games as well, thereby achieving a mutually inferior outcome. This is the same problem surmised to plague the social contract and public goods, but in a state of nature there is no external authority to impose sanctions, through which actors could achieve cooperation.
The sanctions approach to solving a multi-agent, repeating, Prisoner’s Dilemma game thought reflective of numerous social circumstances is limited for being costly and even infeasible or counterproductive to impose.[647] One difficulty is that the agents who are hired as enforcers likely have their own agendas and may be subject to corruption (breaking rules in their favor) and rent seeking (establishing rules in their favor). Evolutionary game theory offered two theoretical results to the ongoing and consequential study of cooperation. First, by modeling indefinitely repeated encounters, researchers added analytic complexity to their understanding of the profound bind organisms are hypothesized to be in: in their struggle for survival and competition for scarce resources, they necessarily optimize against one another on an individualistic and momentary basis. This individualistic competition is believed to be insurmountable because actors straying from this behavioral norm fail to survive and propagate, and it represents a fundamental principle of life best captured by noncooperative game theory.
Second, theorists used the Nash equilibrium of mutual-best-reply and the evolutionary replicator dynamic method to show that in indefinitely repeated games, cooperation can emerge without the introduction of external sanctions through the evolutionary stable strategy of conditional cooperation. This not quite ESS is also called Tit for Tat, or reciprocal altruism. Tit for Tat lies at the heart of neoliberal theory, which argues that even under the most minimalist assumption of myopic strategic self-interest, cooperation can still emerge. This chapter presents the contours and derivation of this behavioral strategy.Neoliberal theorists have unabated enthusiasm for the Tit for Tat strategy because it seems to be the only solution to the endless Prisoner’s Dilemma encounters, perceived to characterize not just human society but life in general, that does not require a strong state. Hence, conditional cooperation appears to be the silver bullet for Prisoner’s Dilemmas. This is apparent in the United Kingdom’s authoritative Stern Review: The Economics of Climate Change (2006). The London School of Economics economist Sir Nicholas Stern provides a comprehensive though succinct overview of state-of-the-art game theoretic findings on cooperation and public goods. His report observes,
Reciprocity plays a key role in situations where the players facing the prisoners’ dilemma have the opportunity to play repeated games and remember the previous choices of the other player. In particular, many players adopt a strategy of conditional cooperation, in which they contribute more to the provision of the public good the more others contribute.[648]
This chapter introduces the Tit for Tat resolution of the repeated Prisoner’s Dilemma game and the hopes, such as Stern’s, pinned on it to solve the vexing problem of cooperation.
However, Tit for Tat more resembles an unstable physical equilibrium, like a ball balanced on the head of a pin, than it does a feasible approach to cooperation.
This is because the conditions to actually tame a strategic rational actor by relying on conditional cooperation are so idealized that they cannot actually sponsor behavior leading to large-scale collaboration.[649] Nor is the approach of reciprocal altruism helpful for actors who either know the end point of their time frame for interaction or place higher emphasis on the present than the future. My points are three. First, conditional cooperators coexist in a purely rarefied theoretical realm presuming indefinitely repeating interactions of two individuals with perfect memory. Second, the efforts necessary to institutionalize the conditions turning Hobbes’s Foole and Hume’s knave into reciprocal altruists are not practically feasible. If reciprocal altruism is to enforce cooperation endogenously, then an elaborate practice of rewards and punishments must emerge to reflect the actors who prevailed in Axelrod’s theoretical experiment, and the environment must perfectly resemble its iterated PD structure. Alternatively, if institutions are built to exogenously facilitate Tit for Tat behavior, then institutional designers must introduce mechanisms to ensure monitoring, transparency, and sanctioning conditions to nudge behavior into that resembling voluntary cooperation. And third, the hope of producing conditional cooperation through indefinite play and perfect transparency, on the premise that all life must obey the laws of individualistic maximization, ultimately fails to even realize that cultural artifacts and sociability could embody a different set of principles of conduct. These differences become evident in any attempt to use Tit for Tat to enforce a conduct of truth telling.[650]