Auctions, Negotiation and Hybrid Mechanisms in Procurement

Principal Investigators: Achim Wambach, Elena Katok

With the advances in information technology, auctions were in widespread use in procurement since the late 1990s. However, in recent years, the use of auctions decreased and hybrid mechanisms, combining features of (standard) auctions and negotiations, have more and more become the norm. These hybrid mechanisms are the focus of our subproject.
In the first funding period we developed a better understanding of buyer-determined reverse auctions (BDRA), where the buyer decides on the winning bid, taking other criteria beside the price like quality into account. We could show that dynamic BDRAs, which play an important role in practice, are prone to tacit collusion. In BDRAs there is usually uncertainty about the buyer’s final decision rule and it is this uncertainty that enables collusive behavior. We also investigated in how far BDRAs lead to a different degree in the level of trust and cooperation between buyers and suppliers than auctions. We found significant differences between the two procurement formats. These results were organized by the application of a model that incorporates other-regarding preferences. We also analyzed in how far favoritism, i.e. the special treatment of a favorite bidder, is possible if non-transparent negotiations are used for procurement instead of auctions with clear rules. We could show that the common perception that auctions are less prone to favoritism does not hold in general, as in auctions the scope of the procurement process has to be defined by the agent in charge, thus allowing for favoritism.
In the second funding period, we will widen the scope of our project by analyzing another commonly used feature of hybrid mechanisms, the “request-for-quote” (RFQ). RFQs potentially provide information about the number of bidders and the costs or cost distribution of bidders. The implications of this information on auction and negotiation design are the contents of our project. In two subprojects the influence of the RFQ on the subsequent negotiation design will be analyzed theoretically, experimentally and with field data. Two other subprojects extend the work done in the first funding period. In cooperation with projects P1, P2 and P5 we will investigate the influence of the negotiation protocol on the degree of trust between buyer and seller. In another subproject, we continue to investigate in how far auctions are more or less prone to favoritism than negotiations, now considering the situation where the agent in charge can bend the rules of the auction (rather than the scope) to favor one bidder.