Cost and quality of service performance in the Chilean water industry: A comparison of stochastic approaches

Abstract

The evaluation of efficiency can be of great value to water companies and regulators to adopt policies and design incentives to enhance performance. This study delves into the implications of employing distinct methodologies, namely the classical Stochastic Frontier Analysis (SFA), Bayesian SFA, and Stochastic non-parametric Envelopment of Data (StoNED), to evaluate cost and quality of service efficiency within the water industry. Chilean water companies reported average efficiencies of 0.623, 0.583, and 0.522 using the SFA, BSFA, and StoNED approaches, respectively. Furthermore, the SFA analysis suggested that the performance of water companies experienced a decline of −0.59% per year from 2010 to 2018. In contrast, the BSFA and StoNED estimations indicated an opposite trend, with annual performance improvements of 0.51% and 0.17% respectively, over the same period. These findings underscore the critical role of selecting appropriate methodologies when interpreting and comparing efficiency results for making informed long-term decisions.