Demand-Driven Innovation and Spatial Competition Over Time Under Heavy-Tailed Signals

Abstract

We study robustness of the model of demand driven innovation and spatial competition over time with log-concavely distributed signals in Jovanovic and Rob (1987) to heavy-tailedness assumptions. We demonstrate that implications of the model remain valid for not extremely heavy-tailed distributions of consumers' signals. However, its properties are reversed in the case of signals with extremely thick-tailed distributions if the firms employ the sample mean of the signals to estimate the ideal product. We further show that conclusions of the model continue to hold under the only assumption of symmetry of signals if a more robust estimator of the ideal product, the sample median, is chosen as the product design.

Keywords: Robustness, heavy-tailed distributions, demand-driven innovation, spatial competition, firm growth, Gibrat's law, signals, investment, information, sample mean, sample median, majorization

JEL Classification: D92, D83, C62