Classical competition and equilibrium : an agent-based analysis

In A Mathematical Formulation of the Ricardian System, Pasinetti (1960) lays out the foundations of what has been dubbed the canonical classical model. He proves the model to be logically consistent and determinate in all its macro-economic features, and derives the solutions for all key variables independently of demand conditions. The model thus provides macroeconomic foundations to the classical theory of distribution. This paper examines the decentralised, competitive mechanism underlying the macroeconomic outcomes. First, we model a classical economy with capitalists, workers, and landlords and define the notion of a Classical Competitive Equilibrium (CCE). A unique CCE exists in a large class of concave classical economies and the resulting income distribution is proved to coincide with that of Pasinetti's canonical classical model. Second, we use an agent-based model in order to examine more explicitly the decentralised competitive mechanisms at play in the classical economy. We show that a realistic competitive interaction between boundedly rational agents with localised knowledge generates classical gravitational dynamics with the key distributive variables oscillating around their equilibrium values.