Endogenous Discounting and Economic Dynamics

We study a discrete-time optimal growth model with endogenous discounting. The discount factor may depend on both consumption and the capital stock, and intertemporal utility is modeled as a discounted sum of instantaneous utilities, with the sum of discount factors equal to one. We show that this specification preserves the invariance of optimal paths and steady states to affine transformations of the instantaneous utility function, providing a general and flexible framework for analyzing economic dynamics under endogenous time preference. We prove that optimal capital paths are monotonic, and steady states depend on initial conditions. We also show the robustness of poverty traps under endogenous discounting: in several examples, for a set of parameters with positive measure, the optimal path converges to a positive steady state only if the initial capital stock exceeds a critical level and otherwise converges to the origin.