International Conference on Economics, Finance & Business, London

HOW CAN PSO AUCTIONS FOR RAIL PASSENGER SERVICES BE MORE COMPETITIVE?

DEJAN TRIFUNOVIC, NIKOLA STOJADINOVIC, BOJAN RISTIC

Abstract:

From 2024, competitive tendering for public service obligation (PSO) railway services replaced direct contract awarding in EU countries, with exceptions in some cases. In reverse auctions for PSO, a bidder with the lowest bid wins the contract and receives the compensation according to the auction rules. Each bidder has a cost of fulfilling the contract, which is his private information, and submits a bid related to his cost. Auctions can be organised for gross contracts, where procurement agencies keep the revenues from ticket sales, and net contracts, where operators collect ticket revenues. In auctions for gross contracts, there is only a private value component (bidders’ cost), and the possible incumbency advantage of a historical operator might only be related to its cost advantage. Auctions for net contracts contain private and common value components (ticket revenue). The primary source of historical operators' incumbency advantage might be superior ticket revenue information. Hence, auctions for gross contracts are a superior solution to level the playing field for competition. In auctions for net contracts, the winner's curse problem induces entrants to bid very cautiously, and they frequently lose auctions against the historical incumbent who has a superior signal on common value and can better avoid the winner's curse problem. If Public Transport Authorities – PTA choose auctions for net contracts, then auctions that enable signal revelation during the auction are preferred, such as hybrid Anglo-Dutch auctions. When entrants obtain significant market shares, collusion between bidders in PSO auctions in the form of a territory division might appear. A first-price auction is significantly better for deterring collusion than an English auction. However, some hybrid auctions, such as the hybrid Amsterdam auction, can even better deter collusion with asymmetric bidders, which is undoubtedly the case in PSO auctions where, in many cases, the historical incumbent is a stronger bidder than the entrant. Based on the data sample from Czech PSO auctions for passenger rail services, and by using collusion screens, we determine that collusion is not currently an issue. The optimal reserve price is the highest compensation per train/km the PTA is willing to pay. This is also a tool for preventing collusion and avoiding unjustifiably high compensations. The PTA might also set the lowest acceptable compensation to avoid the winner’s curse problem in net auctions. Moreover, the lowest acceptable compensation destabilises the cartel agreement by preventing the cartel’s punishment strategy.

Keywords: PSO auctions, incumbency advantage, winner’s curse, collusion