Elements d'economie des reseaux "hub-and-spokes" aeriens. Experience americaine, perspectives europeennes

Elementary economics of air transport hub-and-spokes networks. American experience, european perspectives

• Auteur / Autrice : Bénédicte Molin
• Direction : Alain Bonnafous
• Type : Thèse de doctorat
• Discipline(s) : Sciences économiques
• Date : Soutenance en 1997
• Etablissement(s) : Lyon 2

Mots clés

Mots clés contrôlés

États-Unis

Économie industrielle

Transports aériens

Mots clés libres

Transports aeriens

Etats unis

Economie industrielle

Europe

Résumé

Apres la dereglementation du transport aerien domestique aux etats-unis, les compagnies americaines ont restructure leur exploitation en reseau hub-and-spokes. La logique economique au fondement de cette configuration tient aussi bien aux economies d'echelle et d'envergure presentes dans le transport aerien, qu'au jeu concurrentiel entre les compagnies. La liberalisation du transport aerien intra-europeen requiert la comprehension de ces mecanismes : les reseaux hub-and-spokes peuvent-ils se developper a un niveau intra-europeen, et constituer une structure d'exploitation aussi robuste qu'aux etats-unis ? les compagnies nationales europeennes sont deja basees sur des "hubs naturels", dont le fonctionnement s'apparente plus ou moins a celui des hubs americains. Par ailleurs, l'evolution en cours semble favoriser l'hypothese d'un developpement du hubbing en europe. Un modele de duopole est construit pour tester cette hypothese. Il se presente comme un jeu a deux etapes, en qualite de desserte et en prix, chaque sous-jeu etant en information complete mais imparfaite. Ses resultats donnent les conditions auxquelles un reseau hub-and-spokes serait viable. Ainsi les reseaux hub-and-spokes ne peuvent emerger comme configuration generalisee d'equilibre que si, sur certains marches, l'equilibre est en strategies mixtes. Cette conclusion supporte deux interpretations alternatives mais complementaires. Soit une telle morphologie est par nature instable, soit elle constitue une des meilleures reponses concurrentielles pour chaque firme en presence d'incertitude sur le choix de sa concurrente. Le manque de statistiques empeche l'application du modele a l'europe. Une confrontation a l'observation disponible a neanmoins ete effectuee. Elle teste la pertinence du modele en usant du critere popperien de falsifiabilite. Elle utilise aussi le modele pour apprecier qualitativement les perspectives europeennes en matiere de configuration de reseau.

After domestic air transportation was deregulated in the united states, the american air carriers have turned their networks into hub-and-spoke systems. The american case study reveals that the economic mechanisms at the basis of such systems are the scale and scope economies typical of air transport, as well as the competition between the airlines. The liberalisation of intra-european air transport requires the full understanding of these mechanisms : are intra-european hub-and-spoke networks liable to develop and be as strong a configuration as it is in the usa? all the european national air carriers are already based at a "natural hub", but the operations there do not always look like those at the american hubs. On the other hand, the current evolution seems to favour the hypothesis of the growth of hubbing within europe. A duopoly model is built to test this hypothesis. It is a two-stage game, in routing pattern and price, each sub-game being of complete but imperfect information. It produces conditions, in terms of origin-destination traffic and market length, at which a hub-and-spoke network should be economically viable. These conditions also show that such networks are liable to constitute a generalised configuration at equilibrium only if the equilibrium on some markets is a mixed-strategy equilibrium. This result can be interpreted in two different but complementary ways. This configuration may either be fundamentally unstable, or it can be one of the best replies of each firm when faced with uncertainty about its rival's choice. The lack of adequate statistics prevents any application of the model to europe. Its results have nevertheless been compared with the available data. Owing to this comparison, the relevance of the model has been tested, using popper's criterion of falsifiability. The model was also used as a framework in which the european perspectives of future network configurations have been qualitatively appreciated.