Theoretical evidence of existing pitfalls in measuring hubbing practices in airline networks

Abstract

This paper tries to assess the error committed when measuring an airline's hubbing degree through spatial concentration indicators, such as Gini, Theil, coefficient of variation or even the Hirschmann-Herfindahl. As long as the "spatial concentration" does not seem to explain by itself the main features of the "hubbing" networks, we show that its use presents many shortcomings and may lead to senseless results, most of them related to the use of aggregated data instead of a more detailed routes' distribution through each node, wherein true connecting behavior of passengers precisely lies. Hence, our final consideration of a high concentrated airport as a "hub" or a mere "technical base" is the result of real connections and not only potential or speculative connections. In the previous literature, it is often remarked that H-S has increased the concentration on the specialized hub airports. However, we show that connection and concentration are two different concepts, and that this apparent closest relationship may be blurred for some network configurations.