Professional Documents
Culture Documents
By Vanessa Kamp
Based on currently available studies, Methodology and
German airports appear to be less Data
financially efficient and productive Data Envelopment
compared to other airports in Europe Analysis measures the
and in particular to non-European air- relative efficiency
ports. The reason for this might be the according to Farrell
high degree of vertical integration at (1957). It is a non-para-
German airports. It is important to metric approach that
note though, that only a small number uses linear program-
of international studies have included ming to construct a
German airports; a national study on piece-wise linear fron-
measuring the technical efficiency tier, which is determined by the effi-
does not exist so far. cient airports of the sample (see figure DEA can either focus on input mini-
1). The concept of measuring the tech- mization with constant outputs, or it
This case study measures the techni- nical efficiency with linear program- can calculate an output maximization
cal efficiency of sixteen international ming was first introduced by Charnes, model by holding the inputs constant.
airports in Germany from 1998-2004 Cooper and Rhodes (1978). The decision is very often up to the
with Data Envelopment Analysis An advantage of the DEA is that it can management to affect certain vari-
(DEA) and creates a ranking of the handle multiple inputs and outputs ables. Furthermore, when applying
selected airports. It is one of many to within a single analysis without any DEA, one has to assume either con-
measure the overall efficiency of difficulties of aggregation. Instead of stant returns to scale (CRS) or vari-
German airports as part of the weighting factor quantities as is done able returns to scale (VRS). This
research project German Airport when measuring the Total Factor depends on whether all airports can
Performance (GAP). This paper deals Productivity (TFP), DEA optimizes operate at an optimal scale. If not, it
with the analysis of technical and traf- the weights with linear programming: is more appropriate to assume vari-
fic data, since adequate financial data where θk’ indicates the efficiency able returns to scale, because it
is not yet available. score of every airport k’. zk are the decomposes the technical efficiency
weights that are determined by the score into a) scale inefficiency and b)
Figure 1: Data Envelopment Analysis optimization process. A ‘pure’ technical efficiency.
value of θk’=1 indicates a
point on the frontier and Only a single output has been used, as
thus a technically efficient this case study was also part of an
airport. This linear pro- analysis with Stochastic Frontier
gramming problem must Analysis (SFA), which needs the func-
be solved K times, once tional form of a production frontier.
for each airport, hence θk’ Here, the author decided to choose the
has to be obtained for annual passenger volume as the output
each firm (Coelli et al for the model. Cologne-Bonn as an
2005). airport with high cargo volume has