Search, Failure, and the Value of Moderate Patience

Abstract

Conventional wisdom suggests that successful innovation in complex and novel environments requires patience – persistence despite failures – to ensure broad search for good solutions. Using an agent-based simulation model, I show that the link between patience and innovation is complex: Moderate levels of patience promote broad and effective search. High levels of patience, in contrast, can have unintended effects and even decrease performance despite increasing the degree of search. Furthermore, because translating the gains of patience into performance improvements requires time, low levels of patience are generally optimal for shorter time frames. My findings indicate that paying attention to how patience affects both the explorative and exploitative aspects of search can show when patience may effectively boost innovation.

Keywords: search, experimentation, innovation, complexity, agent-based simulation

JEL Classification: C63, D83, O31