The objective of this work is to analyze the cases of de-hubbing during period 1997-2009 in the world-wide air transport network. To the best of our knowledge, this is the first attempt to study de-hubbing in a systematic way. In order to identify those cases, this paper firstly addresses the issue of which quantitative conditions must be met for airports to be identified as de-hubbing cases. These conditions include the declining presence of the hub carrier, or hub alliance, within the airport, that results in a decrease in the number and quality of connections offered. The second phase is to study what happens after de-hubbing by clustering the cases into homogenous scenarios. Our results show that, on average, airports that suffered de-hubbing did not recover their original traffic in 5 years. Results suggest that de-hubbing is not likely to be reversible. When hub carriers are replaced at least partially by low-cost carriers, the airports on average show faster recovery trends. The most frequent case is when, after de-hubbing, the airports traffic declines. The impact of de-hubbing on the number of destinations is less severe than its effect on offered seats.
