Research

Abstract: 

This paper studies how viewing corporate cartels as networked activities helps explain their stability. Using a dataset of 1, 300 cartels prosecuted between 1990 and 2017, we analyze cartel duration through graph-theoretic measures of firms’ centrality and network density, and compare two cartel network structures—connected and clustered—to identify which is associated with greater cartel stability. Our identification strategy relies on an instrumental variable, the second-degree firms’ centrality, drawn from the applied graph theory literature. First, doubling in firms’ average centrality is reducing the expected cartel duration by 29.04%. Second, a 10 percentage point increase in the share of repeat offenders, used as a measure of network density, is linked to an increase of the expected cartel duration by nearly 12.5%. Third, cartels embedded in connected networks are significantly more stable than those in clustered networks, with a 20% higher expected cartel duration. Overall, our findings highlight the importance of accounting for the network dimension of cartels to better understand the stability and recidivism of corporate cartels.