Tricky figures : why the assessment of competition in banking industries is rather a matter of taste

Economic studies on the degree of competition (DC) in banking systems use various measures which are subsumed under the 1) structure- (e.g. Hirschman-Herfindahl index), 2) conduct- (e.g. Boone indicator) or 3) performance-oriented approach (e.g. Lerner index). Yet, the respective empirical operationalizations of the different DC measures are expected to represent one central construct – the true DC of a banking system. We review 35 studies covering 15 European banking systems from 1998 to 2007. Contrasting the central construct hypothesis, we find substantial differences in the produced DC measures. Thus, the economic validity of derived conclusions regarding the competition intensity is chal-lenged.