The Validity of Input Aggregation in DEA Models: A Statistical Test

We address the problem of input aggregation in DEA models within a broader framework and provide a direct link between input aggregation in DEA on the one hand and the test of parameter restrictions implied by input aggregation in an explicitly specified production function on the other. We show that when input prices vary across firms, the DEA LP problems for measuring efficiency scores of individual firms from the aggregated model have to be appropriately modified. An empirical application of the revised model using data from Indian manufacturing sector reveals that the validity of aggregating production and non-production workers into a composite labor input using Banker's F-test cannot be rejected.