3. Self-Command in Adam Smith’s Theory of Moral Sentiments: A Game-Theoretic Reinterpretation

Ortmann and Meardon (1995b) provides an overview of this paper and Ortmann and Meardon (1993, 1995a). The latter two papers bring similar game-theoretic approaches to bear on the evolution of the general rules of morality in The Theory of Moral Sentiments and problems of public goods provision and externalities in The Wealth of Nations.

From modern micro-economic theory (Holmstroem and Tirole 1989; Tirole 1988; Kreps 1990) to modern macro-economic policy (Barro 1990) and the political economy of institutions and decisions (Ostrom 1990).

Passions are sensations, feelings, or emotions—Smith often uses these terms synonymously—that are excited by some external action or occurrence.

Few of the general rules of morality are ever put in writing—they are informal rules learned by similar example, based on an implicit understanding of what is proper and improper. Morality, for the most part, cannot be legislated for the same reason that high quality of goods cannot be legislated-third-party enforcement costs are simply too high. It is the common knowledge and general acceptance of general rules which constitutes the social fabric; without it, society could not exist.

For an interesting argument along these lines, see also Smith (1937 [1776], 669–670 and 674).

Two-sided simultaneous-move intrapersonal prisoner’s dilemma games have been modeled and defended by Kavka (1991, 1993), among others. Modeling the acquisition of self-command as a one-sided simultaneous-move intrapersonal prisoner’s dilemma can be justified theoretically on the same grounds.

Note that for the Man Yesterday, the proper choice is associated with praise­worthiness and the improper choice is associated with blame-worthiness; for the Man Today, the diagonal entries are associated with praise-worthiness and the off-diagonal entries are associated with blame-worthiness. The upper-left cell lies at the inter­ section of praise-worthy choices for both players; this cell represents respect for the general rules of morality. The lower-left cell lies at the intersection of blame-worthy choices for both players and defines the breakdown in social fabric. As Smith tells us, the “very existence of human society” depends upon the game’s outcome gravitating toward the upper-left cell (1982, 163). Such gravitation is the evolutionary bedrock upon which rests the period-by-period calculation in the game of self-command.

A closely related parameterization can be found in Friedman (1991) and Pitchik and Schotter (1987). Their parameterization leads to a mixed strategy equilibrium in one-shot games. If monitoring costs are high and the temptation to provide low quality is strong, the mixed-strategy Nash equilibrium will approach the one to be identified presently.

Meaning that the row player will react to the column player’s choice of “real” by choosing “improper” forever afterward, and similarly the column player will react to the row player’s choice of “improper” by forever choosing “real.” In this way, the column player protects himself from any future damaging (improper, routine) outcomes; he also punishes the row player, denying him the higher payoffs he could earn from (proper, routine) or (improper, routine) outcomes.

Coleman (1990, 1993) uses a similar approach in sociology.

Etzioni (1986). See also Brennan (1989) and Lutz (1993).

For example, Marwell and Ames (1979, 1980, 1981).

See, for example, Hoffman et al. (1993), Smith (1991), Smith and Walker (1993), Harrison (1992), Yezer et al. (forthcoming) and Ortmann and Tichy (1995). See also Sally (1995) for an important meta-study, though he does not address recent methodological developments.