Airline and high-speed rail competition in Europe: towards a comeback of air transport?

Abstract

At a time when the EU attempts to reduce the negative impact of transport on the environment, meeting demand for long distance travel becomes more and more problematic. To address this problem, the EU and several national governments have opted for a revitalization of European railways and the development of new lines for high-speed trains (HSTs) with the hope that mode substitution "from air to rail" will allow to meet demand for long distance travel at a lower environmental cost. At the same time, the liberalization of air transport has led to a dramatic (quantitative and geographical) development of air services and to their increasing affordability, particularly with the boom of low-cost airlines (LCAs). Yet, the latter are changing the terms of competition and complementarity between HSTs and air services. In this context, the aim of our communication is twofold: 1. On the whole, to check the spectacular dynamic of airline services (number of routes and flights) against the growth and the spatial expansion of European HST services (min. 250 kph). 2. On a selection of city-pairs served by both air and HST services, to show that the volume of air supply often decreases in case of HST competition, but not systematically nor definitively because of the re-cent boom of the LCAs. To achieve this, we will use the results of our PhD on the global dynamic of air services, updating them for the very last years (first objective) thanks to the databases of OAG 1991-2010 (describing the scheduled world air supply in an exhaustive and disaggregated way). For the second objective, we will use the OAG datasets, the maps and timetables published by Thomas Cook and additional data from the rail operators. We will thus identify the European city-pairs connected by HSTs fully or mainly moving on HST lines. For these city-pairs, we will analyse the evolution of the volume of air service in terms of number of flights and seats. We will then interpret the results according to variables such as distance, difference in travel time, frequencies, presence of LCAs, etc.