I have a great book by Tony Lane called The Union Makes Us Strong. Its about working class politics and the trade unions as enduring forms of working class self-organisation and self-representation. But there is a greater moral here. Collective organisation and cooperation is about resisting oppression and exploitation, detecting and resisting and eliminating free riding - the blight of the modern world. The video makes the point very succinctly. But Robert Axelrod and games theory also emphasises how we can build cooperation into our affairs, increase proximity, sustain interaction, generate reciprocity, build trust, encourage cooperation, discourage defection, detect and punish free riders, and ensure that we have the actions and behaviours which give exhortations to the common good some real social substance. Moral exhortation (and appeals to scientific fact for that matter) imply a social identity in which the individual good and the social good are connected. That identity is not available in a market society, it has to be created. The upshot is, rather than appeals to change behaviour from the outside, we change behaviour from the inside. Its all about increasing interactions and generating reciprocity through creating social proximity, encouraging cooperation and eliminating free riders. (Why do we let the big emitters steal our future? We have detected them, we cannot punish them within prevailing relations. Theres no such thing as a free lunch. The capital system is one big free lunch for some, and we need to organise and reclaim control of our common affairs to put an end to this. Cooperation, social proximity, interaction, reciprocity. Union is strength. Anyway, the video fits what Ive been writing this week (rough and needs polishing, like my head). The prisoner’s dilemma involves a one-off encounter between two individuals. In such a scenario, egoism rather than altruism is the rational choice for each individual. However, as Axelrod argues, most real life interactions are repeated and ongoing rather than one-off, isolated occurrences. We meet many people over and again in a social environment. The Iterated Prisoners’ Dilemma describes how the prisoners’ options change once the one-off scenario is repeated over time. Axelrod organised a tournament and invited people to enter a programme based on an iterated Prisoners’ Dilemma. A second round of the tournament was held to reinforce the results. The clear winner in both tournaments was ‘TIT FOR TAT’, a programme with the following strategy: On the first interaction with a program—cooperate. On each successive move after that—do whatever the opponent did last time. A simple strategy of reciprocity that gathered more points overall than other, more subtle, aggressive and highly designed programs. Most interesting of all is that TIT FOR TAT gained the most points overall without beating any programme in particular. Whilst no program scored fewer points against TIT FOR TAT, they scored much fewer points overall against each other, so that TIT FOR TAT emerged victorious. Another significant discovery was that whilst a population of defectors is immune to ‘invasion’ by individual TIT FOR TATs, clusters of TIT FOR TATs could invade such a group so long as they have at least a small proportion of their interactions with each other. At the same time, TIT FOR TAT is immune to invasion, not just from individual defectors but from clusters of defectors. Finally, Axelrod demonstrates that the mechanism of TIT FOR TAT possesses a ‘ratchet’, going on to give historical and biological examples. Once present as a cluster, TIT FOR TAT can proliferate and become dominant. Axelrod closes with these enduring words: ‘The key to doing well lies not in overcoming others, but in eliciting their cooperation’ (Axelrod, 1984, p. 190). TIT-FOR-TAT’s success lies in its being nice, retaliatory, forgiving and clear. Being nice prevents it from getting into unnecessary trouble. Being retaliatory discourages the other side from persisting whenever defection is tried. Being forgiving helps restore mutual co-operation. Being clear makes it intelligible to the other, eliciting long-term co-operation. (Axelrod, 1984, p. 54) However, TIT FOR TAT is only a successful strategy in mixed groups. A group of players using the strategy between themselves will tend to gain a higher overall pay-off than other players who are constantly cheating on each other. TIT FOR TAT protects itself by withdrawing cooperation from any player who defects or attempts to free ride. Denied the cooperation of the well-off group, the pay-off of the free rider or defector will suffer, encouraging them to adopt the TIT FOR TAT strategy themselves. The players act to maximise their own pay-off, there is no demand that they be altruistic or act out of a consideration of the overall good. The system does, however, require trust, even if it does not actually need to be expressed. If free riders go without punishment, free riding will become the most productive long-term strategy, undermining the whole system. The detection and punishment of free riding by the withdrawal of cooperation is key. The one-off prisoner’s dilemma is self-destructive in that the most rational thing for each player to do is defect, thus bringing about the least optimal outcome for all. But when the situation is repeated over time, each player stands to benefit by attempting to establish cooperation with other players. This has nothing to do with an ethic of altruism changing the values of the players and everything to do with changing the game so that it is rational for each to seek to cooperate with others. In an iterated situation, a player’s cooperation is remembered and returned, a virtuous cycle of reciprocity leading to the cooperative society. Axelrod establishes a couple of technical conditions for this outcome: ‘the two key requisites for cooperation to thrive are that the cooperation be based on reciprocity, and that the shadow of the future is important enough to make this reciprocity stable’ (Axelrod, 1984, p. 173). From this follows a need for appropriate scale and social proximity. TIT FOR TAT works best in relatively small groups. To anticipate future actions and built trust between players requires the number of interactions in a given amount of time be maximised, thus building expectations of meeting players in the near future, giving reason to cooperate in expectation of reciprocal action. Conversely, the larger the group, the less interactions there will be in a fixed amount of time, the lower the expectations of meeting players again in the near future, the less trust there will be, the less reason there will be to cooperate. Ultimately, as greater numbers reduce interactions and dilute transactions, we return to the one off Prisoner’s Dilemma, where the most rational thing to do is to defect. For this reason, the Iterated Prisoner’s Dilemma requires a group of such size and scale as to enable the players to meet each other again and regularly, building trust relations and expectations, making it rational to cooperate. There is an equality between the players, so that they have equal power to cooperate or defect, with identical pay-offs for the choices they make. There is no marginal utility differentiation that would induce the players to value their pay-offs differently. Where there is differentiation, a player may be so powerful as to be able to defect against another, even in the expectation of meeting again. Since defection from the less powerful will not affect the powerful player, there is no motivation for the powerful player to establish cooperation. The case for equality savours a little of the principle Rousseau lays down in The Social Contract: I have already defined civil liberty; by equality, we should understand, not that the degrees of power and riches are to be absolutely identical for everybody; but that power shall never be great enough for violence, and shall always be exercised by virtue of rank and law; and that, in respect of riches, no citizen shall ever be wealthy enough to buy another, and none poor enough to be forced to sell himself: which implies, on the part of the great, moderation in goods and position, and, on the side of the common sort, moderation in avarice and covetousness. Rousseau The Social Contract Book II ch 2 There must be transparency, players having access to what was done last time they met others, encouraging players to reciprocate. Transparency is essential for the detection and punishment of free riders. Niceness. TIT FOR TAT must begin with cooperation in order to achieve mass cooperation. Defection the first time it interacts with a new player initiates a cycle of reciprocal defection that ends in mass defection. Freedom. TIT FOR TAT’s decisions to cooperate or defect require no extraneous rules or motivating influences, beyond what a partner did last time they met. The satisfaction of these requirements means that all players can be sure that their decisions to cooperate or to defect will always be met in kind, establishing the conditions for an evolutionarily stable strategy.
Posted on: Tue, 21 Oct 2014 12:55:31 +0000
Trending Topics
Recently Viewed Topics
© 2015