Dynamic efficiency and the two-part golden rule with heterogeneous agents

This paper studies the role of the two-part golden rule as a demarcation line between efficient and inefficient steady states in the neoclassical two-generations-overlapping model with heterogeneous agents. If agents differ regarding their labor endowment, the golden rule ceases to serve its watershed role. Except for the case where all agents’ Engel-curves are linear and identically sloped lines through the origin, some agents’ maximum utility is always associated with a capital intensity that exceeds (or falls short of) the golden rule level. If heterogeneity is introduced on the preference side, the golden rule is never optimal for all agents. These results stem from an intra-generational redistribution of resources which is entailed by the competitive markets.