Searcharxiv⌕ Search

arXiv subjects

Isabela Villamil

Publications and source records attributed to Isabela Villamil.

2 recordsLinked to original sources

Recurrent losers: A network approach to cover-bidding patterns in public procurement

We develop and apply a network-based method to examine patterns consistent with cover bidding within procurement markets, using only bidder participation and winner identities. A statistical test identifies relationships in which firms win unusually often when particular competitors also bid. These relationships form a directed graph with links from recurrent losers to recurrent winners. We use maximum $k$-core to classify possible organizational forms of cover bidding and propose tender exposure -- the share of contested tenders involving a reported relationship -- as a market-level indicator of potential cover-bidding risk. Applied to 200,065 contracts procured by district engineering offices of the Philippine Department of Public Works and Highways (DPWH) from 2017 to 2025, covering three congressional terms, the method identifies 8,728 directed relationships across 407 of 564 markets. The $k$-core classes differ in network breadth, relationship depth, and bidder roles. In simpler networks, each recurrent loser is usually tied to a single winner, while highly interconnected networks involve multiple recurrent winners and broader groups of bidders that rarely win. Some of the deepest relationships and highest levels of tender exposure occur in intermediate networks, showing how relationships can reach much of a market without extensive interconnectedness. Offices named in official investigations more often contain highly interconnected networks than other offices. Two of the three offices referred for bid rigging are among the twelve in our ranking with the highest exposure in 2023-2025.

physics.soc-ph↗

Computational Approaches to the Study of Corruption

Studying corruption presents unique challenges. Recent work in the spirit of computational social science exploits newly available data and methods to give a fresh perspective on this important topic. In this chapter we highlight some of these works, describing how they provide insights into classic social scientific questions about the structure and dynamics of corruption in society from micro to macro scales. We argue that corruption is fruitfully understood as a collective action problem that happens between embedded people and organizations. Computational methods like network science and agent-based modeling can give insights into such situations. We also present various (big) data sources that have been exploited to study corruption. We conclude by highlighting work in adjacent fields, for instance on the problems of collusion, tax evasion, organized crime, and the darkweb, and promising avenues for future work.

physics.soc-ph↗