On the structural determinants of corruption
Ivan AymalievWhile economists have studied the micro and macro-level factors for the variation in individuals’ perceptions of corruption, they have not examined its hidden social order. This article addresses three interrelated questions on the stochastic processes in corruption networks: What network topology smooths consumption in corruption? How do corrupt ties form? What prior and concomitant ties follow corrupt relations? We draw upon Diego Gambetta and Federico Varese’s theoretical framework of corruption and test it using triad census, conditional uniform graph tests, OLS network regressions and exponential random graph models on a cross-sectional dataset comprising forty police corruption networks (N = 693, ties = 1,360) across the globe in the period 1957-2019. Precise images and measures of network structure are reconstructed from reports of police (il)legal relationships in over 1,500 historical documents. Regularities in the structural patterns across the networks are traced using cluster analysis.