Nature and Causes of Economic Fluctuations: Evidence from the U.S. Based on a Nonlinear Autoregressive Integrated Process

Business cycles and economic growth have long been studied separately, hindering understanding of the nature and causes of economic fluctuations and growth. Here, we present an economic model that incorporates both deterministic trends and persistent fluctuations, derived from a general economic data generation process obtained using the dimensionless nonlinear autoregressive integrated (NLARI) process. To explore the nature and causes of economic fluctuations and growth, we develop a unified test to identify whether the economic data generation mechanism is stable fixed points, unit roots, cyclical oscillations, or chaos. Empirical studies show that as the economic resilience coefficient changes from strong to weak, the data-generating mechanism for U.S. economic indicators shifts from a stable fixed point (dynamic equilibrium) in the supply-side economy to an unstable unit root (divergence) in the demand-side economy involving income distribution. Therefore, the U.S. economy is in partial dynamic equilibrium of the supply-side economy. The low resilience of the demand-side economy allows for the accumulation of noise from exogenous shocks, leading to persistent fluctuations. We find that the economic deterministic trend is ([Formula: see text]/[Formula: see text])[Formula: see text], i.e., time [Formula: see text] multiplied by the ratio of the mean of exogenous supply, demand, and technological progress investment shocks to the economic resistance coefficient. A general dynamic equilibrium economy maximizes growth from investment by avoiding [Formula: see text] growth losses from permanently persistent fluctuations. Thus, ([Formula: see text]/[Formula: see text])[Formula: see text] is the maximum value obtained along the maximum profit growth at the von Neumann equilibrium price. The investment path that creates the maximum ratio [Formula: see text]/[Formula: see text] is the optimal growth path.