Corruption dynamics in public procurement: A longitudinal network analysis of local construction contracts in Guatemala

Spending concentration, political influence, and collusion violate rules and principles of open and fair public procurement, leading to corrupt contract allocation. This study adopts stochastic actor-oriented models to test the evolution of these forms of procurement corruption risks in a longitudinal network study of 33579 construction contracts pertaining to Guatemalan local governments from 2012 to 2020. We identify a range of network configurations, based on past empirical research and theory, that capture different patterns of suspicious micro tendencies suggestive of corruption. We show how these micro tendencies shift in strength according to changes in electoral cycles and anti-corruption interventions, thus shedding light on how interventions may temporarily impact corrupt behavior, and how it may adapt and persist after a period of transition. The results indicate that collusion and spending concentration play significant roles in sustaining the risk of corrupt contract allocation, and that this behavior is able to rebound even after the introduction of anti-corruption interventions and new political regimes. The findings underscore the importance of local interventions and advocate for network approaches to enhance transparency, accountability, and long-term anti-corruption efforts.