Standard modeling procedures treat the behavior of economic variables using stochastic-determinate equations. Two alternatives to this approach arose in the last quarter of the twentieth century: deterministic nonlinear equation models and agent-based models. The class of deterministic nonlinear equation models produce dynamic behavior without exogenous shocks, but these models are populated by agents incapable of self-organizing. Agent-based models, on the other hand, allow us to study interactions between artificial actors and the phenomena that emerge from these interactions. In so far as economic phenomena — like prices, firms, nation states and business cycles — are products of human action but not human design, agent-based models are a more productive way forward in formal modeling of complex adaptive systems such as a market economy.