Analysis of collusive bidding behaviour

Researchers have attempted to develop methods that detect collusive bidding. But no method can detect collusion with certainty unless it is based on legal evidence. A method is proposed to detect collusive bidding behaviour that improves the performance of previous methods. It analyses the historical bidding data provided by a construction owner in a two-step approach which is mainly based on a multiple regression model. The first step involves identifying the potential cartel bidders using the residual test and the cost structure stability test developed in earlier work. The second step is the focus of this paper and involves comparing the behaviour of the potential cartel bidders and non-cartel bidders by analysing bid distributions, their cost dispersion, and the differences in their cost structures. After conducting the second step of the study, it was found that the suspected cartel bidders identified in Step 1 behaved in ways to confirm collusion. Also, in an unrelated search, it was found that two of the six potential cartel bidders who were identified in this study had been audited by the public agency for bid fraud, and that another potential cartel bidder had been found guilty by the courts and forbidden from doing business with the public agency.