Game Theory and the Social Contract, Vol. 2: Just Playing (Economic Learning and Social Evolution)

Rating:
Summary: Important Contribution to Political Philosophy
Review: Binmore treats ethics not as a system of rules justified by Reason, but as by contrast, ethics the scientific study of how humans behave and think. Binmore reports extensively on contemporary ethnographies of hunter-gatherer societies, believing that such societies mirror the social and material conditions the human race faced during its formative period as a species. Such societies have no division of labor except for gender, and are politically egalitarian, decision-making power being quite equally distributed among the adult males of the community. Binmore infers that fairness norms must be self-enforcing, and cannot depend on a hierarchical leader (a "philosopher-king") to enforce ethical principals. Moreover, since a division of labor (except for gender) is absent, deliberations in such groups approximate the `original position.'

Binmore thus offers us a "coevolution of genes and culture" in which the acceptance of original position moral arguments is written into our genes, but the cultural content depends on local environmental conditions and random variation. Again drawing on the ethnographic literature, Binmore focuses on food sharing as the most important rule of justice to be decided by a foraging group. In foraging societies, high variance foodstuffs such as meat are equally shared, irrespective of who made the kill. Equal sharing is thus a moral rule justified by reasoning from the original position of hunters who do not know exactly which among them will be lucky or skilled.

Binmore uses evolutionary game theory to analyze social interactions. This adds a welcome degree of clarity to ethical reasoning. Indeed, Binmore is quite clear that all of his substantive results depend on the plausibility of the game theoretic models he presents and analyzes.

While fairness norms are biologically determined for Binmore, the players in Binmore's games are rational self-interested agents. Thus all of the results of two-person game theory based on the rational actor model can be deployed in analyzing social justice. It follows in particular that "[i]n a well-ordered society, each citizen honors the social contract because it is in his own self-interest to do so, provided that enough of his fellow citizens do the same." (5) There is no sense in which moral behavior is opposed to self-interested behavior. Moreover, since players do not behave ethically in bargaining, there is no sense in which the institutions resulting from their bargaining have any abstract normative standing. "Evolutionists simply seek to understand," says Binmore, "why some types of human organization survive better than others.... evolutionary ethics offers no authority whatsoever to those who wish to claim that some moral systems are somehow intrinsically superior to others.' (179)

Different societies can thus embrace different institutions because comparisons in the original position depend on `empathetic preferences' that are culturally specific. It is in part for this reason that Binmore calls himself a `whig,' by which he means a moderate progressive, not seduced by the grand visions of a totally alternative society as proposed by the Left and the Right. The latter two, he claims, make social judgments in a universal, ahistorical manner that have nothing to do with the actual fairness processes in real societies.

Just Playing is an important and welcome contribution to the literature. The book does, however, have some faults. The most salient is that crucial analytical material and discursive asides jumbled together. One must read the whole book, and make numerous references back and forth, to understand the basic argument. Moreover, the book is intended for a general audience interested in political philosophy, yet even professional economists will find the analytical parts difficult to follow.

Another problem is that Binmore uses evolutionary game theory where it suits him, but abandons it when it does not. For instance, while Binmore uses naturalism to justify the assertion that Homo sapiens is genetically programed to accept the original position, but he gives no empirical evidence that this is in fact the case. Moreover, it is implausible that evolution imprinted us with an original position orientation, but in no other way affected our moral behavior, so that the assumption of Homo economicus remains valid for bargaining purposes. Laboratory experiments reveal forms of prosocial behavior (e.g., rejecting `unfair' offers in an ultimatum game, or punishing free riders in a public goods game) that relate directly to questions of justice and fairness, yet contradict the Homo economicus model. The notion that human sociality can be explained by `enlightened self-interest,' even when accompanied by respect for the original position, will not likely survive a close study of the evidence (See my book Game Theory Evolving, Princeton University Press, 2000).

Rating:
Summary: Upgrading Rawls' "Theory of Justice"
Review: In his exciting theory of the social contract Ken Binmore takes up the discussion that took place in the 70ies after the publication of John Rawls' "Theory of Justice". While he sticks to the idea of a social contract reached through voluntary agreement in the Original Position, he also considers the utilitarian critique such as Harsanyi's. But Binmore does much more than that. He translates Rawls' metaphysical idea of a reflective equilibrium into a two-stage bargaining game with flesh and bones. He stresses the tautological character of game-theoretic tools which in this context becomes an advantage. By comparison of the ethical properties of allocations reached via competitive markets and those reached through bargaining in the original position he tries to identify a demarcation line for the decentralized aggregation of individual preferences. Binmore's book is going to be a challenge to any reader interested in the problem of explaining progress in human societies.