A Machine Learning Approach for Flagging Incomplete Bid-Rigging Cartels
Open Access
- 11 September 2022
- journal article
- research article
- Published by Springer Nature in Computational Economics
- Vol. 62 (4) , 1669-1720
Abstract
We propose a detection method for flagging bid-rigging cartels, particularly useful when cartels are incomplete. Our approach combines screens, i.e., statistics derived from the distribution of bids in a tender, with machine learning to predict the probability of collusion. As a methodological innovation, we calculate such screens for all possible subgroups of three or four bids within a tender and use summary statistics like the mean, median, maximum, and minimum of each screen as predictors in the machine learning algorithm. This approach tackles the issue that competitive bids in incomplete cartels distort the statistical signals produced by bid rigging and it outperforms previously suggested methods in applications to incomplete cartels based on empirical data from Switzerland.Keywords
Funding Information
- University of Fribourg
This publication has 39 references indexed in Scilit:
- Testing for collusion in asymmetric first-price auctionsInternational Journal of Industrial Organization, 2013
- Libor manipulation?Journal of Banking & Finance, 2012
- A Study of the Internal Organization of a Bidding CartelAmerican Economic Review, 2010
- Endogenous cartel formation with heterogeneous firmsThe RAND Journal of Economics, 2010
- Bidding rings and the winner's curseThe RAND Journal of Economics, 2008
- Super LearnerStatistical Applications in Genetics and Molecular Biology, 2007
- Deciding Between Competition and CollusionReview of Economics and Statistics, 2003
- Ohio School Milk Markets: An Analysis of BiddingThe RAND Journal of Economics, 1999
- Detection of Bid Rigging in Procurement AuctionsJournal of Political Economy, 1993
- COUNTERSPECULATION, AUCTIONS, AND COMPETITIVE SEALED TENDERSThe Journal of Finance, 1961