اداره امور و تئوری بازی : چه وقت مزایده های فرانشیز شرکت ها را وادار به پیشنهاد زیادتر می کند ؟

The use of auctions as an instrument of public policy has been hailed as evidence of the utility and validity of game theory. In this paper, we focus on extreme cases – centrally, spectrum auctions in the UK and Germany in 2000 – to argue that the canonical game theoretic interpretations of firm behaviour in some (highly “successful”) auction outcomes are inadequate, that the economics of governance critique of franchise bidding can be extended to provide a better interpretation, and that under specified conditions, notably high uncertainty, policy-makers should design franchise auctions to avoid overbidding, rather than attend to the more conventional challenge of underbidding.

Auctions have become the conventional choice of governments seeking to assign public assets or rights to firms for commercial exploitation, especially for the assignment of spectrum for commercial wireless communications services. The success of these auctions – “success” measured most often by the public revenue generated – has been taken as a clinching demonstration of the validity and utility of the game theoretic foundations of the auction mechanism. Following the adage that the “study of extreme instances often provides important leads to the essentials of the situation” (Behavioral Sciences Subpanel, quoted in Williamson, 1976, p. 75), this paper re-examines the two most spectacular of the so-called third generation (3G) or UMTS spectrum auctions held in Europe in 2000–2001: the British auction of April 2000 and the German auction of August 2000. The argument is that firm behaviour in these auctions eludes interpretation under the canonical game theoretic framework, that this outcome can be better understood under an extension of the economics of governance critique of franchise bidding, and that this revised understanding has policy implications. In the second section, we summarize European 3G licensing, and in particular the British and German auctions, and the economic and political aftermath of these auctions. In the third section, we attempt to estimate the order of magnitude of spectrum value loss experienced by the winning firms. In Section 4, we consider the game theoretic interpretation of the auction results and identify overbidding, in this and other cases, as a challenge to such theory. In Section 5, we offer an alternative approach to franchise auctions based on the economics of governance. Section 6 examines alternatives to the fixed cash-maximand auction mechanism and other possible palliatives to overbidding. Section 7 concludes.

In conclusion, under high uncertainty canonical auction theory imputes an implausible hyper-rationality to economic agents. When public assets essential to entry or sustained presence in a market are to be auctioned, high uncertainty and potential or actual asset specificity are indicators of potential overbidding which should attract the attention of policy-makers. Focus on information asymmetries may be a diversion from the fundamental problem of insufficiency of information. Maximization of auction pricing is not a proxy for maximization of collective welfare nor a unique indicator of the efficiency of public policy. Bidding firms, especially incumbents, are not risk-neutral, but highly averse to the risk of losing the franchise at stake (or, as in the German case, experiencing a more competitive market structure). Public asset auctions in complex policy environments create incomplete contracts with significant implications for collective welfare. In consequence they must be designed with the efficient and effective management of sustained bilateral dependency between firm and state in mind.