One should also mention the disadvantages of the DEA method, which is
characterised by the results’ high sensitivity to atypical data in objects
recognised as models. If the model object is an atypical one, the results of the
analysis of the efficiency of other objects are considerably less credible. One
may also notice a negative impact on the test results of surprising and unstable
results in the case of a strong correlation and linear relationships within the
results, within the expenditures or between the results and the expenditures. The
disadvantage of this method is also the redundancy of the number of efficient
objects, especially in its traditional versions, and a poorly-developed theory of
nonlinear relationships between the expenditures and the results. Another
disadvantage may be the relative nature of the object’s efficiency. In the DEA
method, efficiency is determined against the background of other objects. As
a result, an object with a relatively low efficiency may be considered fully
efficient because the other objects are worse. The opposite scenario is also
possible