TY - JOUR
T1 - The beauty of Dutch
T2 - Bidding behavior in combinatorial first-price procurement auctions
AU - Paulsen, Per
AU - Bichler, Martin
AU - Kokott, Gian Marco
N1 - Publisher Copyright:
© 2020 Elsevier B.V.
PY - 2021/6/1
Y1 - 2021/6/1
N2 - Ex-post split-award auctions are a frequently used form of combinatorial auction mechanism in practice. The procurement quantity is split into several shares and suppliers can submit bids on separate shares as well as on the entire quantity. Markets with diseconomies of scale are wide-spread, but strategically challenging. In a game-theoretical equilibrium analysis, Kokott et al. (2019) have recently shown that in contrast to single-object auctions, there is no strategic equivalence between first-price sealed-bid (FPSB) and Dutch combinatorial auctions. The FPSB auctions are characterized by efficient and inefficient equilibria while the Dutch auctions only possess efficient equilibria. We report the results of extensive laboratory experiments and show that the theory explains the bid data surprisingly well. Importantly, a compound Dutch auction format weakly outperforms the wide-spread combinatorial first-price sealed-bid auction in efficiency and total procurement costs. The results provide guidance for procurement managers in the field.
AB - Ex-post split-award auctions are a frequently used form of combinatorial auction mechanism in practice. The procurement quantity is split into several shares and suppliers can submit bids on separate shares as well as on the entire quantity. Markets with diseconomies of scale are wide-spread, but strategically challenging. In a game-theoretical equilibrium analysis, Kokott et al. (2019) have recently shown that in contrast to single-object auctions, there is no strategic equivalence between first-price sealed-bid (FPSB) and Dutch combinatorial auctions. The FPSB auctions are characterized by efficient and inefficient equilibria while the Dutch auctions only possess efficient equilibria. We report the results of extensive laboratory experiments and show that the theory explains the bid data surprisingly well. Importantly, a compound Dutch auction format weakly outperforms the wide-spread combinatorial first-price sealed-bid auction in efficiency and total procurement costs. The results provide guidance for procurement managers in the field.
KW - Lab experiments
KW - Split-award auctions
UR - http://www.scopus.com/inward/record.url?scp=85093945987&partnerID=8YFLogxK
U2 - 10.1016/j.ejor.2020.09.048
DO - 10.1016/j.ejor.2020.09.048
M3 - Article
AN - SCOPUS:85093945987
SN - 0377-2217
VL - 291
SP - 711
EP - 721
JO - European Journal of Operational Research
JF - European Journal of Operational Research
IS - 2
ER -