رفع ابهام از بهره وری هزینه خدمات آب و فاضلاب ارائه شده اشتراکی

Providing operators with objective incentives for cost efficiency and continuous improvement in the provision of public services are major concerns for regulators. Measuring efficiency empirically is complex and this complexity is accentuated when the same operator is responsible for delivering more than one service (e.g. in order to explore potential economies of scope). Based on a sample of operators that provide water and wastewater services, this paper uses a shared input data envelopment analysis model to measure separately the efficiency of each service. The results show that a single measure may not provide enough information for monitoring multi-utilities. Together with other indicators, the proposed model can assist decision-makers in prioritizing efforts to improve overall efficiency.

Measuring the cost efficiency in the delivery of public utility services is of crucial importance. For the same level of service, higher efficiencies should lessen the burden on rate and/or taxpayers (if there is regulatory pressure). However, when the same operator delivers more than one service, performance measurement becomes more challenging (Torres and Morrison, 2006) and global efficiency measures tend to be less useful. Traditional methodologies do not always highlight in which service efficiency is lower: a key issue for both decision-makers and regulators. In a given territory, water and wastewater services are often jointly provided by the same operator. In fact, empirical evidence supports the argument that there are economies of scope between drinking water supply and wastewater collection/treatment/disposal, especially in smaller utilities (Abbott and Cohen, 2009). Most methodologies used in the literature to evaluate the performance of water utilities only estimate overall efficiencies and do not assess the cost efficiency of each activity (e.g. see Gómez and Rubio, 2008; Romano and Guerrini, 2011 for a general overview of the literature). It could be the case that, for example, a given operator is cost efficient in drinking water supply and inefficient in the delivery of wastewater services. Using an overall efficiency score would not highlight this conclusion in a straightforward manner. Although evaluating the overall efficiency of operators in these cases still has significant value, managing to separate the efficiency of the water and wastewater services could be of further use for decision-makers and regulators. Several methodologies have been used to assess the performance of water and wastewater services (Berg and Marques, 2011).1 A conventional classification is the division between parametric and nonparametric methodologies, and both have their strengths and limitations (for a more detailed discussion see Fried et al., 2008). Despite being widely used in the literature, parametric methodologies require an a priori definition of the cost or production function and the acceptance of various assumptions derived from economic theory (which may reduce the acceptability of the results by some members of the scientific community). Nonparametric methodologies use the information ‘within the data’ to estimate efficiency scores and they do not require as many assumptions or constraints.2 Among the many methodologies available, the data envelopment analysis (DEA) is the most frequently used by researchers. By means of linear programming, DEA estimates a best practice frontier using the inputs and outputs of all observations and computes efficiencies using the most favorable weights for each decision-making unit (DMU). The information asymmetries between regulators (independent agencies or local authorities) and operators (public or private) hinder the effectiveness of the regulatory framework (Berg, 2000). Frequently, the lack of transparency and sufficient detail in the annual statements of the operators do not allow the proper design and monitoring of incentives for cost efficiency. Although there are several operators that already do this explicitly in their financial statements, incurred costs (operations and capital) and staff are not typically allocated to the corresponding service (in our case, water and wastewater). To the best of our knowledge, this is the first application of a nonparametric model designed to estimate the cost efficiency of each output (i.e. each service) in the water sector, when both services are jointly provided by the same operator.3 The objective of this paper is to propose a model for estimating not only the overall efficiency of water utilities, but also the cost efficiency in each of the services provided. In this case, the two services under analysis are drinking water and wastewater services. Using a shared input DEA methodology (see Beasley, 1995; Cook and Green, 2004; Cook et al., 2000; Rogge and De Jaeger, 2012), the authors are also able to report estimates for the cost shares that correspond to each service. Naturally, this methodology could prove to be very useful for regulators and decision makers who wish to benchmark their services against the best practices of the sector. This article is organized in the following manner. After this introduction, Section 2 briefly describes the importance of economic regulation in the water sector. It addresses some international experiences and the difficulties of putting in place an effective framework of incentives for cost efficiency. Section 3 presents the shared input DEA model along with the data used to assess its usefulness (consisting of 253 observations from 45 Portuguese water utilities for the period 2002–2008). Section 4 summarizes the results obtained and, finally, Section 5 provides a discussion, concluding the paper.

The results show that the major share of the total costs of multi-utilities providing water and wastewater services are allocated to drinking water supply (around 64%, on average). The shared input DEA model also allowed us to conclude that there is no statistical significant difference between the efficiencies of drinking water services and wastewater services. Furthermore, it seems that vertically integrated operators (providing ‘retail’ and ‘wholesale’ drinking water and wastewater services) have higher cost efficiencies for both services and therefore also in overall terms. This is an interesting finding with potential policy implications (that go against recent reforms in Portugal). However, detecting the presence of economies of vertical integration is neither the major objective of this paper nor the main usefulness of the shared input DEA model. Additional research should be carried out on this topic regarding the Portuguese ‘retail’ water and wastewater markets. As we have shown (for instance in Fig. 5), used in conjugation with ‘traditional’ methodologies for measuring global performance, the model proposed in this paper to disentangle the cost efficiency of water and wastewater services can generate more information that is useful for the missions of regulators and managers. Since it allows for identifying asymmetric performances in water and wastewater services when these are jointly provided by the same operator, the methodology would especially useful for the operators that are far away from the central tendency (regression), as exemplified in Fig. 3. The usual lack of transparency and detailed financial information regarding the management of water utilities often hinders the effectiveness of regulation. The proposed methodology tries to cope with this reality and provides a solution to the classic problem of information asymmetry. Moreover, as the primary “regulators” of the water and wastewater services provided in their jurisdictions, municipalities are in need of useful tools for monitoring their utilities. Indeed, the exercise carried out in this paper could be repeated individually for all water utilities. The shared input DEA model does not only provide local governments with their ‘global’ picture, it also allows for the identification of those activities for which there is still room for improvements. Finally, the fact that the methodology implemented computes the most favorable cost efficiency score for each operator (as in ‘traditional DEA’ and given the proper constraints), increases the acceptability of the model as a performance assessment tool.