"In this paper I want to begin with the neoclassical supply and demand model of markets (SDM), whose static equilibrium consequences predicted outcomes far more accurately than were anticipated in laboratory experimental tests of the theory actuated by Jevons (1862, 1871; Smith, 1962). The observed predictive accuracy of SDM was not anticipated because complete information on supply and demand was widely believed, thought and taught to be a necessary condition for finding equilibrium. 1 Jevons’ model required him to have complete information in any particular market, as he only articulated a model of market optimal outcomes, and no model of how individual actors might discover those outcomes. Only private information on their own unit values (costs) was provided to the subject-agents in the experiments; yet convergence was rapid. Market participants functioning under principles of motion, known to no one, were finding the equilibrium by means that were no part of the theory, nor any part of their own intentions and awareness."
Smith, V. L. (2018). Causal versus consequential motives in mental models of agent social and economic action: Experiments, and the neoclassical diversion in economics. ESI Working Paper 18-11. Retrieved from https://digitalcommons.chapman.edu/esi_working_papers/249