Ferreira, DiogoCaldas, PauloVarela, MiguelMarques, Rui Cunha2024-04-042024-04-042023http://hdl.handle.net/10400.5/30527There are several ways of aggregating partial performance indicators into composites, each of them with advantages but also shortcomings and caveats. The objectivity demanded for policy-making and regulation of utilities leads many researchers to build their analyses over the so-called Benefit-of-Doubt (BoD), which, in turn, results from the well-known Data Envelopment Analysis (DEA). In a compensatory manner, these models construct piecewise linear frontiers containing the benchmarks, but they disregard increasing marginal products and the existence of non-concavity regions, thus underestimating efficiency. Multiplicative approaches have been proposed to solve these problems; one of these is the geometric (rather than linear) aggregation of variables. But still they fail to solve some problems like the correct treatment of undesirable variables, the existence of regulatory constraints, and the existence of imperfect knowledge of data. Therefore, this paper builds upon a geometric aggregation of performance indicators and proposes some strategies to solve the aforementioned shortcomings of the existing models. The new framework is exemplified and tested with the Portuguese urban solid waste management utilities.engComposite indicators; Multiplicative Directional Benefit of Doubt; Imperfect Knowledge of Data; Regulation; Solid Waste Managementjournal article10.1016/j.eswa.2023.119540