Many feminist economists have argued that neoclassical economics is woefully inadequate for inquiry into the economic lives of women. The neoclassical emphasis on "choice" is misleading, they say: neoclassical accounts tend to pay too little attention (1) to systematic differences in the options available to men and women and (2) to the socially constructed nature of the "preferences" guiding choices. And neoclassical competitive-market-equilibrium stories misrepresent the institutional settings and processes that generate occupational and distributional outcomes: we need much richer accounts with a broader range of actors and of sites and sorts of interaction.(1)
Institutional economists have been advancing similar criticisms for decades. This paper seeks to stimulate discussion of the commonalities between feminist and institutionalist methodological concerns; it focuses upon assessment of one possible alternative analytical approach--game- theory--as applied to one set of phenomena, gender relations in the household.(2)
The Neoclassical Unitary Household
Neoclassical work on many phenomena takes "the household" as the decision-making unit and portrays it as seeking to maximize a unitary household utility function. The analyst usually does not explain how this function is related to the individual utility functions of household members; or Becker's [1974, 1981a, 1981b] "benevolent dictatorship" formulation may be cited, in which the "household's" utility function is that of an altruistic household head.
The stories told within this analytic framework have often served to rationalize gender inequality. Add assumptions about differing relative productivities of men and women in market and nonmarket work, and one gets demonstrations that the assignment of work in the home to women, and women's resulting disadvantaged position on the labor market, are rational and welfare-maximizing.
Both neoclassical and Marxist feminists have expressed the desire for alternative accounts that show the household to be an arena of conflict and contestation, as well as caring and cooperation, and that see primary responsibility for child care and housework as socially imposed on women, not "freely" chosen. There has been an interesting convergence on the part of many neoclassical and Marxist feminists toward game-theoretic approaches, which, advocates argue, neglect neither social structure nor human agency.(3)
Game-Theoretic Modelling of the Household
Bargaining models begin by specifying (1) the object(s) of bargaining; (2) the players' objectives; (3) the set of feasible outcomes and their associated "payoffs" for each player; and (4) the rules by which the outcome is to be determined.
Household members might be seen as bargaining over: the division of tasks; overall labor time and leisure; and the distribution of consumption goods and services (this list could be expanded). We might portray the players as seeking to maximize individual utility.
Regarding the rules for determining outcomes, game theory offers two very different approaches.
In noncooperative games (such as the Prisoners' Dilemma), the players cannot communicate or make binding agreements. Each must choose her actions without being able to coordinate choices with the other. The game protocols that must be specified include (1) the options available to each player at each step; (2) what determines when the game will end; (3) what each player knows about the other's objectives, options, and knowledge; and (4) the expected payoffs associated with each action by each player, for each possible action the other player might take. In the household, a noncooperative game might have each player taking unilateral action regarding work and consumption over a number of "rounds" until a stable (equilibrium) division is achieved.
In cooperative games, in contrast, players can make binding agreements. …