Organizational learning in evolving networks
Building on the classic model of March (1991), we study how alternative network evolution logics (link, distance, random, performance) impact organizational learning and performance. We find that networks formed based on distance (such as homophily) consistently outperform those formed by links (ie rich-gets-richer).
Furthermore, the more attention an organization pays to ‘performance benchmarking’, the poorer its overall performance. Lastly, the continual introduction of new members has very little positive impact on overall performance across all logics.