Humans cooperate a great deal in economic activity, but our two major models of equilibrium – Walrasian competitive in markets and Nash in games – portray us as only non-cooperative. In earlier work, I have proposed a model of cooperative decision making (Kantian optimization); here, I embed Kantian optimization in general equilibrium models and show that ‘Walras-Kant’ equilibria exist and often resolve inefficiencies associated with income taxation, public goods and bads, and non-traditional firm ownership, which typically plague models where agents are Nash optimizers. In four examples, introducing Kantian optimization in one market – often the labor market – suffices to internalize externalities, generating Pareto efficient equilibria in their presence. The scope for efficient decentralization via markets appears to be significantly broadened with cooperative behavior.