(initial caveat: I am not presenting a brief in favor of Sraffa's system; I am going to ignore that system. I think that Joan Robinson's critique (that it was only dealing with comparisons of unrealistic equilibrium points) was sufficient. I don't see it as either a substitute for or a representation of Marxian political economy.) 1. Steve Cullenberg asks Gil Skillman: can the Arrow-Debreu world >>articulate an economic system where a surplus (and hence exploitation) exists as an equilibrium result<< where the economic surplus represents >>a profit opportunity that paradoxically is not being taken advantage of in equilibrium?<< As Gil points out, an economic surplus can be introduced into the A-D Walrasian system in the form of scarcity rents. Henry George's theory of landlord exploitation (based on the unproducibility of land and other natural resources) is a good example here; it could easily be introduced into the A-D system. John Maynard Keynes' theory of rentier exploitation (based on the interest-inelasticity of the supply of saving) might also be introduced into the A-D system (though most of his other insights contradict that system). John Roemer (seemingly without knowing of or caring about his precursors) generalized the scarcity theory of economic surplus by extending it to profits received by those who control the means of production. Unfortunately, such "Roemerian profits" are merely quasi-rents. The existence of profits based on scarcity rents ( "profit opportunit[ies] ... paradoxically ... not being taken advantage of in equilibrium") rather than true rents (in which there is some natural barrier to taking advantage of the profit opportunity). Their existence implies an incentive for capitalists to invest, which undermines the scarcity of means of production and thus the basis for Roemer's theory -- unless we bring in assumptions that Roemer didn't think of. (See Gary Dymski's and my article in ECONOMICS AND PHILOSOPHY, 1991.) I'm afraid that Roemer got swept away by the static and formalistic thinking that pervades Walrasian thinking and never thought about the dynamic implications of his models. 2. There is a very poor match between the "economic surplus" (scarcity rent) which can be introduced the A-D system and Marxian surplus-value. The A-D system is merely a totally unrealistic story of exchange which rejects the concept of realistic (historical) time in favor of merely imaginary (logical) time, while Marxian political economy deals with an economy in which labor is involved in production (which takes historical time), where the products may or may not end up being sold. I argue in a recent article (published in Bill Dugger's book (INEQUALITY: RADICAL INSTITUTIONALIST VIEWS ON RACE, CLASS, GENDER, AND NATION. Greenwood Press, 1996), Marxian surplus-value is more like a _tax_ than a neoclassical scarcity rent. That is, its existence is based on capitalist class power: the reserve army of the unemployed, the control of the labor process by capitalist management, and passive acceptance by the workers. Of course, capitalist exploitation is more than a standard tax system since it involves constant technical change, growing labor productivity, accumulation, and crises. But the point of the rent vs. tax contrast is to discourage people from trying a shot-gun marriage of Marx and Walras (of the sort that Roemer proposes and Gil seems to endorse) not to give a complete picture of Marxian exploitaition. I disagree strongly with Gil's parting words, i.e., that "the Walrasian framework does not rule out any political amd historical processes of determination." Because it is timeless, the Walrasian framework _cannot_ allow for any real processes, except as exogenous. But endogenous to capitalism are "political" processes such as the production process (one site of class struggle), while the structural coercion implied by the reserve army of labor is anathema to Walrasian thought. (If I remember correctly, the existence of Walrasian equilibrium assumes "free disposal" of unused resources, which means that unemployed workers must be shot -- or similar. So Walras requires Pol Pot?) As Gil points out, "wages and interest rates arise in markets for labor power and financial capital." But as anyone who has studied labor economics or money & banking seriously knows, these markets are not Walrasian auction markets operating under perfect information with wages and interest rates determined at the instant of the attainment of equilibrium and not an instant sooner. Further, as Bowles & Gintis point out, both labor-power and financial markets typically involve non-Walrasian rationing and "contested exchange." We have to reject the artificial distinction between politics and economics that pervades Walrasian economics: "economic" sites such as the labor process and the hiring and firing of labor-power involve (micro-level) politics. Further, the dynamics of capitalist accumulation and class struggle (endogenous processes of the conventionally-defined "economy") determine the "distribution of endowments" that Walrasians take as exogenously given. On the other hand, game theory is way cool. in pen-l solidarity, Jim Devine