Rights and Game Strategies – §A

Stephen Boydstun's picture
Submitted by Stephen Boydstun on Wed, 2006-11-29 17:59

Russell Hardin (1982, 1986, 1988) and Robert Sugden (1986) have independently employed game theory to expand our understanding of human rights. The relevance of game theory to rights theory is straightforward if we begin by recalling Ayn Rand’s observation that a right is a moral principle defining and sanctioning an individual’s freedom of action in a social context. Rights are deeply entrenched in social relations. Game theory can inform rights theory precisely because, as Hardin has observed, “to have a right is to be free to take some particular kind of action in the context of actions by others” (Hardin 1986, 69).


The theory of games is the theory of interdependent decision-making. Game theory abstracts the formal, logical properties of social situations in which outcomes depend on decisions of two or more agents. It clarifies the relationship between actions and outcomes in such interactions. When a certain formal game can be recognized in a social situation, all of the properties that game theorists and experimentalists have discovered for the game can be expected in that social situation. Formal games have been used to classify and analyze many types of social interaction. Examples are nuclear deterrence and arms races, processes of bargaining and pricing, and the problem of public goods.


In the general outline of games having two participants, or players, each with only two alternative actions, each player chooses between two alternatives with no way to be sure what the other player is choosing until later. The game is now over. There are four possible outcomes, each corresponding to a particular pair of choices. The payoffs to each player from the four possible outcomes could vary in desirability as follows: One player might prefer outcome A over B over C over D; the other player might prefer outcome B over C over A, with D being perceptibly neither better nor worse than C. Each such dual set of preferences specifies a distinct game.


A. Conventional Character of Rights


Rights are principles of justice, and they are principles of social coordination. Conventions are principles of social coordination. Game theory illuminates both rights and conventions.


Long before the advent of game theory, David Hume concluded that principles of justice are purely conventional rules. In Hume’s view, justice is a convention. Hume did not mean to suggest that principles of justice are arbitrary rules on account of their ultimate conventionality. Rules such as general abstinence from taking the possessions of others are, in Hume’s view, rationally commendable conventions because they are conducive to public utility and our own individual well-being.


Hardin and Sugden find aspects of Hume’s social ideas attractive. They have extended and more solidly grounded Hume’s ideas using game theory. They have each emphasized different elements of Hume’s thought in constructing their own social theories. Hardin and Sugden evidently do not know of one another’s work; neither ever mentions the other.


The theories of Hardin and Sugden regarding conventions build upon earlier discussions by David Lewis (1969) and Thomas Schelling (1960). Hardin and Sugden have each recognized the vitality of one of Schelling’s insights into the emergence of conventions: in order to become a convention, a rule must possess some psychological prominence, some salience, that people might independently select.


Lewis advanced understanding of conventions using Schelling’s coordination games. In these games, there is no immediate conflict of interests; whatever the choices made by the players, it is the case that if either would have been better off had she alone chosen differently than she did choose, the other player would also have been better off. Two motorists approaching each other on a road just wide enough for dual passage face this kind of coordination problem. A convention such as “always drive on the right” can coordinate actions with mutual benefit on such an occasion.


Hardin and Sugden have each followed Lewis’s lead, but have undertaken to show how conventions might emerge in situations containing some incentives to cooperate, but also some immediate conflicts of interest. Coordinated outcomes are possible not only in games of pure coordination, but in games of partial conflict. Hardin concentrates on embodiments of the game of partial conflict known as the Prisoner’s Dilemma (one player prefers outcome A over B over C over D, whereas the other player prefers D over B over C over A). Sugden focuses mostly on games of Chicken (one player prefers A over B over C over D, whereas the other player prefers C over B over A over D) or close relatives of Chicken.


To illustrate the emergence of conventions in a situation of partial conflict, Sugden analyzes the case of two drivers approaching a crossroads. If both drivers were to maintain speed, the consequences would be worse for both than if both were to slow down. It would be best for each driver if only the other were to slow down. The result to be expected most frequently, the natural outcome of the game (#68 in Rapoport, Guyer, and Gordon 1976), in the absence of any convention, is that both drivers will slow down. The game has two equilibriums; these are the two outcomes in which one driver maintains speed and the other slows down. Those are the choice combinations with respect to which either player would suffer if only he or she were to change strategy.


Sugden is an economist with the feathers of Mandeville and Smith. He delights in showing how conventions can arise from the choices of individuals pursuing their own interests with no concern for resulting social patterns. Unfortunately he misses the pertinence of the fact that many individuals are so concerned. Those concerns can provide the salience required for a convention. Wagoners who must cross single-lane bridges might very well be attracted to the rule “yield to an opposing wagon that has already entered the bridge” out of concern for elementary social utility.


The same sort of flaw appears in Sugden’s account of property rights as conventions. His account is stunning nonetheless. He has demonstrated that property conventions would tend to arise that favor one of the equilibrium outcomes in each of recurrent game situations people would face under social anomie, or “the state of nature.” From social conditions in which each would simply take whatever he could lay his hands on, rules for property appropriation would emerge spontaneously that favor either precedent possessors or challengers of precedent possessors.


Why, then, does common law favor precedent possessors? What gives possession the psychological prominence required to be selected as the conventional rule? Sugden suggests several possibilities: There is the saliency for humans of the number one among all counting numbers and of firstness more generally. On average, possessors may attach a higher value to disputed resources and so may be expected to fight harder, or possession may, on average, confer some advantage in a fight.


Sugden recognizes that the expenditure of labor on objects is also commonly perceived as conferring ownership, but he supposes that the saliency of labor expenditure arises derivatively from the saliency of possession. Expenditure of labor is a way of registering a claim of possession, a way of establishing association between person and object that may be seen as a possessory relation. Possibly so, but surely labor could acquire saliency in addition from considerations of simple social utility and feelings of empathy. Social utility should also be admitted as a plausible reason for convergence upon property rules that favor possessors rather than challengers of possessors.


(This article continues under two more subtitles: “Moral Aspect of Rights” and “Utility of Rights.” I will try to display these in separate posts to the general forum. The article shown here is excerpted from one written in 1988 and published in 1994 in Nomos: Studies in Spontaneous Order. I would like to thank the editor Carol B. Low for improvements of expression.)

( categories: )