prior probability

In our previous two posts, we identified a blind spot in the work of John Rawls and Bob Nozick: the problem of betrayal–what game theorists call “defection.” In brief, Rawls’s original agreement might embody timeless principles of justice, and Nozick’s private protection groups might be the first step towards his political utopia, but in both cases, there is nothing stopping people from repudiating their mutual promises or double-crossing each other. What is the way forward, then?

Game theory offers some possible solutions to this trust dilemma. Even in the case of the deservedly famous Prisoner’s Dilemma, where fear of the dreaded “Sucker’s Payoff” leads the prisoners to betray each other (see below), Robert Axelrod and the late Bill Hamilton identified the conditions in which cooperation might yet still be possible: iteration (i.e. playing the game more than once) and uncertainty as to when the last round of the game will occur. But there is also a major blind spot with Axelrod and Hamilton’s work and with game theory generally. In addition to our two prisoners, game theorists silently assume a third player: a judge or strongman who is able to distribute the payoffs of the prisoner’s game and enforce its rules. But who is this secret third actor or deus ex machina? Where does he come from? Why should the prisoners pay him any heed?

In short, overcoming the problem of betrayal/defection–getting cooperation off the ground–appears to be a truly intractable problem, and yet we still see many examples of cooperation and mutual trust around us. Is there a better way of explaining the puzzle of cooperation? Is there a new way forward, one that does not rely on unenforceable original agreements (Rawls), or on ahistorical mutual protection agreements (Nozick), or on secret third-party payoff enforcers (Axelrod & Hamilton and all game theorists generally?) We are certainly open to suggestions …

Where is Player C, i.e. who distributes these payoffs and enforces the rules?

In our previous post, we identified a weakness in Rawls’s theory of justice: the possibility of betrayal once the veil of ignorance is lifted. Now, let’s turn to Robert Nozick’s classic tome Anarchy, State, and Utopia (1974), a book Nozick wrote in response to Rawls’s work. In Chapter 1 of Anarchy, State, and Utopia, Nozick postulates the emergence of voluntary private protection agencies in an ahistorical state of nature. In other words, instead of a hypothetical original agreement à la Rawls, Nozick imagines mafia-like protection rackets filling the legal and political void (p. 12): “Groups of individuals may form mutual-protection associations: all will answer the call of any member for defense or for the enforcement of his rights.” (As an aside, Nozick even retrodicts (p. 13) that “some entrepreneurs will go into the business of selling protective services.”) But there is a blind spot in Nozick’s theory of the pre-political state, the same blind spot that Rawls overlooks in his work: the possibility of betrayal.

In summary, since the private protection services and protection agencies in Nozick’s state of nature are the result of voluntary agreements, who enforces them? After all, Nozick’s imaginary protection agencies are operating in a lawless world, so these private protection agreements are not legally or morally binding in any meaningful sense. And without any external or credible enforcement mechanisms in a pre-political or stateless world, why should we assume that people in Nozick’s state of nature won’t betray each other when it suits them. That is, since Nozick’s private protection agencies are operating in a lawless and stateless world, what is stopping any member of any given private protection agency from repudiating his membership or from betraying a fellow member or from joining another rival protective association?

In a word: nothing.

At one point, Nozick briefly contemplates the possibility of a “breakup” (p. 13) of a private protective association due to internal discord or dissension, but he does not consider the systematic possibility of betrayal or breach among the members of his protection agencies. Nor does he explain why such private agencies would be immune to the problem of betrayal. Yet, we would expect betrayal to occur frequently in a state of nature. To sum up, then, Nozick–like Rawls–ignores the possibility of betrayal. Specifically, just as Rawls assumes that the hypothetical parties to his hypothetical “original agreement” won’t betray each other once the veil is lifted, Nozick too assumes that the private members of his private protection agencies won’t betray each other as well. But in a world where there is no penalty for breach–whether it be Rawls’s hypothetical post-original position or Nozick’s ahistorical state of nature–, these are false and fool-hardy assumptions. The original agreement or agreements for private protection are not self-enforcing.

For the next three posts, we are going to offer a critique of John Rawls’ theory of justice (part 1), followed by a critique of Robert Nozick’s theory of the pre-political state (part 2), and then offer our own alternative to their influential theories. Let’s start with Rawls, shall we? In his book A Theory of Justice (1971), Rawls presents an ingenious thought experiment consisting of a hypothetical or make-believe world in which hypothetical rational actors negotiate a hypothetical original agreement. In this make-believe world, people negotiate an “original agreement” from a hypothetical starting point, i.e. from the original position, an initial position of perfect equality. In addition, these negotiations occur from behind a veil of ignorance in which “… no one knows his place in society, his class position or social status, nor does anyone know his fortune in the distribution of natural assets and abilities, his intelligence, strength, and the like” (1971, p. 208). Now, here’s the payoff: according to Rawls, the outcome of these hypothetical negotiations will be consistent with the principles of justice.

There is a fundamental flaw, however, in Rawls’s elegant thought experiment. Simply put, what happens when we lift the veil? What’s stopping any of the parties to the original agreement from betraying the other parties to the contract or from repudiating any promises made in the original position?

For his part, Rawls states (ibid., p. 208) that the principles of justice agreed to in the original position “are the result of a fair agreement or bargain.” But with all due respect, so what? After all, there are no penalties for failure to abide by the original agreement once the veil of ignorance is lifted. Rawls’s response to this objection is to emphasize (p. 209) that “the obligations they recognize [in the original agreement] [are] self-imposed,” but Rawls is utterly unpersuasive on this point. After all, what difference does it make that the obligations agreed to in the original agreement are self-imposed if there is no penalty for breach? Once the veil is lifted, there is nothing stopping the parties to the original agreement from going back on their word. Worse yet, instead of providing penalties for breach or a credible external enforcement mechanism, Rawls simply assumes (p. 208) that “… having chosen a conception of justice [in the original position], we can suppose that they are to choose a constitution and a legislature to enact laws, and so on, all in accordance with the principles of justice initially agreed upon.” But Rawls is wrong. In a world where the possibility of betrayal is lurking in the background–our world–, we cannot assume this at all.