• Embed Doc
  • Readcast
  • Collections
  • CommentGo Back
 
The Perspective through Gini
The Gini index, when compared tomore conventional indices of concen-tration such as the Herfindahl Index,has distinguished itself for being par-ticularly useful when measuring traf-fic distributions inside airline net-works. One of its salient features isthat it reacts quite well to changes inall parts of its population rather thanonly its most important ones. The reg-ularly observed unequal, log-like dis-tribution of airline networks is quitewell accommodated by Gini.Although one may argue about the proper units of analysis when usingthe index (some tend to favor passen-gers or number of aircraft movements,others choose available seat capacity),the adequacy of Gini for measuringtraffic concentration across airports isnot challenged. In that sense, compar-ing spatial concentration in Europeagainst that of the US may be a worth-while exercise. The analysis might be performed more properly when avail-able-seat-miles (ASM) is used as theunit of analysis to account for generaldifferences in distances within bothgeographical areas.Highlighting such geopolitical differ-ences has far-reaching implications ona micro, airport planning level, as wellas on planning for multi-modal con-nectivity alternatives, or, on a macro- political level, when evaluating therisks of bilateral “Open Skies” agree-ments. Interpreting Gini from such ageopolitical point of view, however,requires decomposing the index fur-ther into finer layers of traffic distri- butions that are particularly relevantto eachpolicy perspective. The mostobvious dimensions that come tomind are domestic routes (whichwould correspond to intra-state routesin the US), routes between the differ-ent EU member states against trafficconnecting states in the US, and long-distance or intercontinental on bothcontinents.These first results alreadyshow striking differences.We find, in the first table, that consid-erably fewerUS airports serve intra-state air traffic as compared to EUdomestic (within the nation state). Thedifference in concentrations for suchtraffic is not statistically significant between both geographical areas. Thisempirical finding confirms the lack of sensitivity of Gini to the size or dis-tance distributions between territoriesfor each geographical area. However,when factoring in distance(see ASM),sensitivity becomes more important:with a 90 percent confidence interval,we can conclude that such traffic isspatially more concentrated under the
NETWORK RESEARCH
e-zine edition, Issue 38
1
Spatial Concentration in EuropeVersus the US
From Gini Analysis to Geopolitical Assessment 
Liberalization of air traffic in both the US and Europe has resulted in distinct patterns of spatial concen-tration of traffic distribution on each of the continents. Statistical indices, such as the Gini coefficient,make it possible to measure inequalities in this distribution across airports. Applying this comprehensivemeasure of traffic to each common market (EU, US) helps to assess today’s results of deregulation and toplan airline networks and airports for the future. The same approach can be applied to long-distance orintercontinental flights. The aim of this paper is to account for geopolitically induced differences thatcan be stressed when comparing traffic distributions across continents. Our research findings showimportant and potentially far-reaching asymmetries between (Western) Europe and the US for most of the geo-political dimensions. The implications of our analysis for policymaking are outlined in part 3 of this contribution.By Dr. Hans Huber
Legend: AS =available seats, ASM = available seat miles, CI = confidence inter 
val
Table 1:
Spatial concentration in Europe versus the US for different route-types
 
influence of longer intra-state routesin the US. We find that the number of airports serving such routes is some43 percent higher in the US. On a 95 percent confidence interval, we findsuch traffic significantly more con-centrated in the US in comparison toEurope. Although distance suggestslittle influence on such traffic con-centration in Europe, again it adds tospatial concentration in the US. Theresults for long-distance/interconti-nental routes are also surprising:although the US shows some 38 per-cent less airports serving long-dis-tance/ intercontinental routes, con-centration is significantly higher (90 percent CI) in Europe.Also, we foundwith Gini that the distribution of suchroutes across the different airports isthe most even of all, given an averagevalue of 0.745 for AS.These findingsimply a strong national approachwithin the European ‘Union’. Asexpected, distance does not stress thedifference in concentration between both geographical areas due to thenature of the route. One importantconclusion is that European airporthubs depend much more on long-dis-tance/intercontinental routes, whereasspatial concentration in the US seemsinduced internally (domestically),and is able to take advantage of, onaverage, longer distances.
Clustering Airline Networks intoStrategic Groups
The economic agents that determinethese traffic distributions are, of course, the various airline operators.The networks that they operate differ from each other – grouping themalong network features [1]into strate-gic groups is an intermediate step thatshows differences in market structure between Europe and the US.When examining the second table indetail, we find a much more fragment-ed market structure in Europe, withvery small networks. At the other extreme, the largest US networks aresignificantly bigger than the Europeanones. There are more very large net-works in the US: the largest 7 carrierseasily outperform the largest 6European in terms of monthly depar-tures from their busiest hubs. Thenumber of average routes per AP issignificantly higher,and these US car-riers show significantly steeper slopes(which indicate a higher spatial con-centration) in traffic distribution aswell. The case around AmericanAirlines shows significantly morelong-distance links than any EU- based airline. More airlinesoperatingsmaller networks in the 7 to 25 APrange can be found in Europe, butthey show a lower tendency towardsspatial concentration. The same num- ber of mid-size carrier networks can be found in the US and Europe (thereare 17 of them in either geographicalarea): European ones show more con-nections, on average, per airport (4.5versus 3.2in the US), although theEuropean ones tend to be lessspatial-ly concentrated at their main airports(showing departures in a range of log(2.79 to 2.96) versus a log(3.25) inthe US). A final distinctive feature for these 17 mid-sized airline networks isthat the US groups show almost nolong-distance links as compared tosignificant intercontinental route serv-ice in the EU.
Decomposition of Route-Typetraffic Distribution alongStrategic Groups
In the final step of our analysis, we cannow highlight how different strategicgroups in both geographical areasmanage traffic for distinct route types,i.e. intra-state/ domestic versus inter-state/ intra-European versus intercon-tinental. Our findings (detailed statisti-cal results can be obtained from theauthor upon request) can be summa-rized as follows.The US geopolitical environmentallowed for the emergence of nationalor even regional hubs that cater to theoperations of mostly large,high densi-ty networks. They serve larger geo-graphic distances on averageand their respective market coverage shows lit-tle overlap with airports that are beingcovered by other strategic groups.Intra-state traffic contributes to higher spatial concentration, which is proba- bly enhanced by distance advantagesin a common market.One may arguethat this configuration is likely to beadopted in Europe as well, given thelonger period for which the US mar-ket has been de-regulated. On theother side, the underlying geo-politi-caldrivers in Europe (for example:national policies by Member States,etc.) are not likely to change soon and
e-zine edition, Issue 38
2
Table 2:
Clustering of airline networks into strategic groups
of 00

Leave a Comment

You must be to leave a comment.
Submit
Characters: ...
You must be to leave a comment.
Submit
Characters: ...