The issue of input/output slacks in validating data envelopment analysis-based Malmquist productivity index: an example from banking

Data envelopment analysis (DEA) models have been used in formulating the Malmquist index to measure productivity change over time periods. In this article, we question the validity of this index in the presence of input/output slacks that appear in DEA models used. We demonstrate with an application to banking that Malmquist index loses its meaning whenever slacks are present. As a corrective method, to neutralise the effect of slacks, we propose intrinsic assurance regions to be appended to the DEA models.