Religion, Politics and Luck: Explaining the Rise of Sustained Innovation

e institutional and path-dependency approaches used to explain innovation in Britain during the Industrial Revolution generate prediction errors when applied elsewhere. Here, innovation is modeled as a dynamic coordination game with multiple equilibria. Tests of the resulting propositions based on 116 successful inventions between 1700 and 1849 suggest that there were two distinct categories of innovations. Cooperative innovations (those involving two ormore unrelated principals) required religious or linguistic signals indicating willingness to cooperate. Noncooperative innovations (those involving a single inventor) were more the result of chance. Sustained innovation of both types was favored by revolutionary political change.