This paper was my attempt to build a theory of "public factors of production" such as scientific knowledge and to see whether there is any way that Arrow-Debreu general equilibrium theory can cope with the nonconvexities that arise. I propose that the economy might be partitioned into provision of "social overhead goods" and ordinary goods in such a way that given the quantities of social overhead goods, the remaining activities satisfy the assumptions needed for the existence of competitive equilibrium. I proposed and examined a notion of "Lindahl-Hotelling equilibrium" in which individuals were assigned Lindahl prices for the social overhead goods. These Lindahl prices depend on the effects of the social overhead goods on competitive prices for ordinary goods. This allows a nice application of duality and indirect utility functions.