Evaluating the efficiency of decision making units using panel Data Envelopment Analysis – empirical example for Polish commercial banks

The author makes an attempt to assess the relative efficiency of Polish commercial banks using non-paramteric Data Envelopment Analysis. In the empirical research, use has been made of panel data – 24 commercial banks in 4 years (2001–2004). The results of sequential DEA method (which examines 24 banks individually in each year) are compared to the results of integral DEA method (which examines 24 banks in four years at the same time – one bank appears in such analysis four times, i.e. there is 24 × 4 = 96 objects under examination). Owing to this approach it is easier to reveal which unit is truly efficient and which one is only accidentally recognized as efficient. The author comes to the conclusion that the integral DEA method is more reliable, because the results are more stable and fewer units are recognized (incorrectly) as efficient.