1 Introduction

The demand for ample quantities of high-quality water for various purposes is consistently increasing. For this reason, integrated management of water resources that considers the following aspects is required: population, economic development, water pollution, uncertainties in the face of climate change, resource availability, environmental concerns and sustainable development (Santamarta et al. 2022a, b). Within Europe, the Water Framework Directive (WFD) (EU 2000) serves as the overarching legislation encompassing all pertinent water management domains. The WFD goal is to attain a favourable status for water bodies, impacting their quality not only in terms of the advancement of aquatic ecosystems but also concerning water for human consumption and its utilization in diverse economic activities (Gallego-Ayala and Gómez-Limón 2009). Hence, as stipulated in Article 9 of the WFD, the principle of cost recovery for water-related services must be considered, including environmental and resource-related services, in accordance with the ‘Polluter Pays’ principle and a pricing policy that encourages efficient use of water resources by all users (Hughes and Malmqvist 2005). Water pricing is contemplated in the WFD with a dual role: a) as an incentive for efficient use; and b) as an instrument for cost recovery, clearly establishing that all member states must develop and implement water pricing policies that enable the recovery of the cost of water in all its dimensions, i.e., financial, environmental and resource costs (Tsitsifli et al. 2017).

However, the lack of specificity in the directive allows each country to develop its own regulations. Taking advantage of the generality of the WFD complicates decision making and transparency (Galioto et al. 2013). For this reason, according to Pellegrini et al. (2023), the WFD would benefit from having a greater variety of tools to address local specificities from various perspectives, as well as considering quotas, subsidies, and stakeholder participation as key elements of the process.

Focusing on the information about what constitutes a water service, it is any activity developed by an agent for the benefit of a user (domestic, industrial, agricultural and/or public) in relation to water resources (Mylopoulos and Fafoutis 2012). These services are subject to recovery through tariffs and charges, or as self-service payments. They are of two types: i) abstraction, impoundment, storage, treatment and distribution of surface water and groundwater; and ii) collection and treatment of discharged water. Depending on their nature within the water cycle, they can be either upstream services, where one agent makes water available to another who, not being the final consumer, will again offer another service; or downstream services, where the agents make water available to the final user (Fig. 1).

Fig. 1
figure 1

Water services from catchment to end-user and reuse

Full size image

The financial costs of water services must be estimated through information from the agent providing the service or the financing agent (Alamanos et al. 2020). Environmental costs are the economic cost of the actions needed to minimize the environmental damage associated exclusively with the provision of water services (Cortignani et al. 2018). They are conceived as a penalty rate for pollution associated with the provision of water services. These are the costs of the measures that correct and/or avoid deterioration in water bodies due to the provision of a service (Kanakoudis et al. 2012). Moreover, the Program of Measures is the instrument that coordinates and integrates the measures necessary to achieve the environmental objectives set by the WFD (Carvalho et al. 2019). Finally, resource cost is considered the cost of scarcity, understood as the cost of the opportunities renounced when a scarce resource is allocated to one use instead of another (Albiac et al. 2020); it also relates to lost revenues due to the alternative water uses that were not chosen (Turner et al. 2019).

The Canary Islands form a Spanish archipelago and represent an outermost European region that encounters distinct challenges in water management compared to continental areas (Fig. 2). The climate is subtropical, and temperatures are strongly linked to the effect of the trade winds on these islands. The relationship of the Canary Islands with water is a very close one, where the exploitation of groundwater resources, through horizontal water mines (galleries) and coastal wells, stands out (Santamarta et al. 2014). Of lesser importance is the contribution of surface water, since precipitation is of the order of 350 mm per year (González-Morales and Ramón-Ojeda 2019). Seawater desalination is a solution that is gaining more and more ground in the archipelago (Gómez-Gotor et al. 2018), due to high population and tourism, as well as agriculture, which has the highest water demand of all the sectors in the Canary Islands (Díaz et al. 2011).

Fig. 2
figure 2

Canary Islands (Spain), where each of the seven islands has its own River Basin Management Plan

Full size image

One of the main peculiarities of the Canary Islands is that most of the groundwater is privately owned (Custodio et al. 2016). Private investment to finance drilling works that illuminate groundwater (galleries and wells) was translated into shares. The shareholders, in turn, are grouped into Water Communities or Irrigation Communities, and they can use their water in any way they wish – use their own water, sell the shares or the water to third parties and/or cede the temporary management of the rights of the shares to a third party (Macías Hernández 2009).

According to Khemlani et al. (2021), the Canary Islands face a number of cross-cutting and specific challenges affecting water management. Among the cross-cutting challenges are the following: i) the orography of the terrain (Coello-Bravo et al. 2020), which poses a challenge in terms of access to drinking water in many localities; ii) the increase in the local population; iii) the ever-growing foreign and national tourism (Gundelfinger-Casar and Coto-Millán 2018); and iv) climate change, which is the new paradigm from which the water challenge in the Canary Islands must be approached (Santamarta et al. 2021). Areas of specific interest include droughts, the decrease in groundwater resources due to exploitation of the aquifer since 1920 (Custodio et al. 2016), the low quantity of surface water resources, and the increase in desalination in order to meet the high water demand of the Canary Islands (Schallenberg-Rodríguez et al. 2014).

However, the increase in desalination in the Canary Islands must be viewed in conjunction with the islands' energy consumption (Santamarta et al. 2022a, b). Desalination requires significant amounts of energy, and the Canary archipelago has a very low penetration of renewable energy in its electricity mix (Gobierno de Canarias 2022). As a result, the water cycle in the Canary Islands notably contributes to CO2 emissions into the atmosphere (Cruz-Pérez et al. 2022). In addition, wastewater treatment and management also present important economic and environmental challenges, which have yet to be exploited in the Canary Islands (Ruiz-Rosa et al. 2016).

Another important challenge in the Canary Islands is the renovation of water supply pipes that generate high water losses in the Islands, with consequent loss of the resource (Santamarta et al. 2021). This will need to be alleviated in the near future to avoid water wastage and will require a large investment of public money. As mentioned earlier, the water resource in the Canary Islands comes from groundwater (a valuable resource that has sustained life in the Canary Islands) and desalination. Therefore, losing drinking water in potable water distribution networks means losing a millennia-old groundwater resource, as well as using significant amounts of energy and increasing emissions into the atmosphere to produce that water. This is in addition to the impact of brine on the marine ecosystem in the Canary Islands (Kress et al. 2020).

The stochastic frontier model allows us to assess the efficiency of different water production techniques. Dealing with the comparison of heterogeneous units is common to efficiency studies of water production, and for this reason, many studies have employed the frontier approach (Guerrini et al. 2018; Laureti et al. 2021; Molinos-Senante and Maziotis 2019). Through this model, we estimate probable water yields according to production technology, considering that there is a random variable based on variations found in historical data. This random variable could be explained by the losses due to distribution networks. Considering the island context, we have the advantage of having water production data with different technologies.

The aim of this study is to analyze the cost recovery established by the River Basin Management Plans of the second cycle for the Canary Islands, and to analyze the results for each of the islands, in order to better understand the existent mechanisms that contribute to good status of the water bodies.

2 Methodology

To calculate cost recovery related to the water cycle, it is necessary to identify and describe water uses (upstream services, downstream supply services and downstream sewerage systems), and water services (defense against floods, environmental defense and water management in general). An estimation of the investment costs per year is made. In order to do so, it is necessary to calculate the Equivalent Annual Cost (EAC) (Wang et al. 2018), using the following equation:

$${EAC}_{inv.}=\frac{\left(1+r\right)^n-1}{r\cdot\left(1+r\right)^n}\cdot I$$
(1)

where r is the discount rate (as a decimal), n is the useful life (years) and I is the initial investment (at constant prices).

Consideration should be given to water service grants, investment, capital and accounting costs, as well as administrative, operating, and maintenance expenses. This information can be obtained from public and state agencies (Table 1). With this information, capital and exploitation costs can be estimated by compiling the data.

Table 1 Classification of the sources of information by agencies
Full size table

As mentioned before, environmental costs are the estimated cost of measures taken to reduce environmental impacts associated exclusively with water provision services. Internalized and outstanding environmental costs are obtained using this method. Outstanding costs are estimated by the EAC equation (Eq. 1) for the measures yet to be executed. Internalized costs are the maintenance and operation costs of the actions currently being executed; therefore, for these measures, costs must be deducted from their financial costs. The only actions considered for estimation are aimed at environmental correction of deterioration caused by the provision of water services linked to Article 2.38 of the WFD (EU 2000).

Resource costs are described from market instruments and how they allow improvement in economic conservation of the resource and environmental flows. They signify the cost of scarcity or the lost income from alternative water uses. In this study, they come into consideration during a river transfer or the production of water through desalination. For the case study, the only numerical value considered for cost recovery is that of desalination, as no river transfers are made in the islands. They must be considered as the production cost of the water sale price at the plant.

Water services must have a cost recovery instrument, which is necessary to characterize the provision of water services as an income. Table 2 sums up the main instruments for cost recovery considered for the Canary Islands.

Table 2 Cost recovery instrument by service and agent providing the service
Full size table

The WFD methodology for cost recovery establishes that in order to calculate cost recovery, the costs must be compared with income obtained from users for the provision of the different services (EU 2000). The results are obtained for each agent from public and private entities.

The cost recovery rate is calculated by dividing annual income and the annualized cost for water services. This rate must be specified, at least for urban water supply and agricultural and industrial uses. Evaluation of the penalty for contamination and recovery of environmental costs must also be considered. The cost recovery methodology that has been implemented in the Canary Islands follows the structure shown in Fig. 3.

Fig. 3
figure 3

Process followed in the calculation of cost recovery for the Canary Islands

Full size image

2.1 Econometric Analysis

The methodology used in this section refers to the principle of technical efficiency applied to cost recovery for water-related services. In this section, an analysis is made of the technical efficiency of water services, considering costs related to the production and distribution of water services. The objective is to obtain an estimate of technical inefficiency in the provision of water services, based on available data. The process of water production and distribution, as shown in Table 2, involves both public and private institutions. By estimating technical inefficiency, we are estimating the degree of efficiency/inefficiency of the institutions involved in the production and distribution of water services. However, we should be cautious in the interpretation of the results, due to the limited data available, so that we do not have a panel of data that would allow us to make a better assessment of the differences by islands. Using cross-sectional data, we are also unable to observe changes over time.

The model presented below considers the costs related to upstream water services (surface and groundwater) and desalination on the production side, and the costs of distribution services, such as downstream irrigation water distribution and urban supply on the distribution side. Based on these production and distribution factors, a stochastic frontier is configured for the cost function. The stochastic frontier model is motivated by the theoretical idea that no economic agent can exceed the ideal "frontier" and the deviations from this extreme represent individual inefficiencies (Belotti et al. 2012). The ideal "frontier" means the minimum expenditure required to produce a bundle of outputs, given the prices of the inputs used. Self-service is excluded from the model, as it is not quantitatively representative. Surface water comes from reservoirs, groundwater is abstracted through aquifers, and desalinated water is obtained from desalination facilities. The total costs of water services, which form the dependent variable, are composed of environmental costs, resource costs and financial costs. Environmental costs are conceived as a penalty rate for polluting. Resource costs represent the scarcity of the resource or the opportunity cost of its use. Financial costs are the costs necessary to obtain the resource.

2.1.1 Stochastic Frontier Model

The main purpose of using a stochastic frontier model is to estimate the efficiency of cost recovery of water, considering the production and distribution of water services. The costs of water services are influenced by the technology employed and also by all the inputs necessary to carry out production and distribution. Stochastic frontier analysis is the most popular tool for efficiency analysis. It is a stochastic approximation method to determine production efficiency (Aigner et al. 1977; Battese and Coelli 1992; Belotti et al. 2012; Farrell 1957; Guerrini et al. 2018; Laureti et al. 2021; Molinos-Senante and Maziotis 2019). In our case, the objective is to estimate an optimal cost frontier. A deviation from the frontier reflects both inefficiencies and noise in the data. For this reason, two residual terms are considered in the model: one captures the error from unpredictable (random) disturbances; the other captures technical inefficiency.

The model is as follows:

$${CT}_{i} = {\alpha + X}_{i}\beta + {\varepsilon }_{i}$$
(2)
$${\varepsilon }_{i} = {\nu }_{i} - {s\mu }_{i}$$
(3)

where CTi is the vector of total costs; \({X}_{i}\) are the outputs; s = 1 for production functions; s = -1 for cost functions; \({\varepsilon }_{i}\) is the stochastic disturbance term, where \({\nu }_{i}\) is the random term containing the unpredictable disturbances or noise in the data, which is assumed to have a normal distribution and independent from \({\mu }_{i}\); \({\mu }_{i}\) is nonnegative disturbance, standing for technical inefficiency in the cost inefficiency of the cost function; α and β parameters are estimated where β represents the vector of technology parameters. The inefficiency term is assumed to be a half-normal distribution \(\left[{\mu }_{i} \sim {N}^{+}(0, {\sigma }_{\mu }^{2})\right]\). Each observation i represents each island in our case study. The function defines a minimum level of costs associated with the outputs (production and distribution) that correspond to water services. Two factors affect the efficient production term, and they have to do with the imbalance between marginal costs and marginal revenues. The second factor has to do with technical efficiency, which is analysed in this case. Technical efficiency is understood as the capacity to produce the service at minimum cost in financial, environmental and resource terms. The function defines a minimum cost limit. The difference between this theoretical benchmark and the actual benchmark is known as technical inefficiency. The model is estimated by maximum likelihood (ML).

From the results (Table 3 and 4), it can be seen that all the production and distribution factors considered are significant and have a positive impact on the generation of costs. This result is therefore in line with expectations. The factors with the highest impact on recovery costs are urban water supply (0.39), followed by groundwater services (0.26) and desalination (0.15): \({\varepsilon }_{i}\) is positive, so there is no technical inefficiency problem, although the noise in the data is significant; \({\nu }_{i}\) containing the unpredictable disturbances is significant and somewhat high. The term \({\mu }_{i}\), relating to technical inefficiency, is not significant. Technical inefficiency ranges from -0.02 to 0.03, the values being quite small and including zero in the confidence interval, as can be seen from the λ values in Table 3. More interesting, if the data allowed it, would be to compare technical efficiency by island and over time, which would require a panel of data. It is very likely that a panel of data would reduce the noise in the data, making it possible to establish a ranking of efficiency between islands (Schmidt and Sickles 1984).

Table 3 Values obtained from stochastic frontier model
Full size table
Table 4 Estimating technical inefficiency (Jondrow et al. 1982)
Full size table

In order to separate the error term of the stochastic frontier model into its two components (the inefficiency deviations and the stochastic deviations), Jondrow et al. (1982) proposed a method that allows estimation of the level of technical inefficiency for each observation in the sample, thus overcoming a disadvantage of the stochastic frontier model. The estimation of Jondrow et al. (1982) technical efficiency is based on calculation of the mean of the conditional distribution of the error term, which represents the technical inefficiency (\({\mu }_{i}\)), conditional on the composite error of each productive unit i (\({\varepsilon }_{i}\)). Table 4 shows that the average technical inefficiency is 0.00059, with a standard deviation of 0.00092. This result is consistent with the non-significance of the term \({\mu }_{i}\) that we observe in Table 3.

3 Results and Discussion

The breakdown of cost recovery by cost type for each island, and for the second cycle of the River Basin Management Plan, is illustrated in Table 5. In general, at least part of the financial cost is recovered in all islands. Tenerife and Gran Canaria show the highest values for financial cost recovery. This could be due to the fact that, as these two are the capitals of the two provinces of the Canary Islands, this is where most of the population is located and they have more tourism, which requires more water infrastructure development (Custodio et al. 2016).

Table 5 Classification of the costs in the islands
Full size table

In Tenerife, complete financial cost recovery for desalination and water reuse could compromise its development (CIATF 2018), as it is still necessary to invest in the improvement of these facilities to reach a good quality of water with the least impact to the environment (Dallavalle et al. 2021). El Hierro, La Gomera and La Palma show very low financial cost recovery, due to the fact that a low percentage of the demanded water is billed to users, and information from private services is not always up to date or available (CIALP 2018).

Lanzarote and Fuerteventura are highly dependent on desalination as a source for urban supply, due to the low availability of surface or groundwater resources (Gómez-Gotor et al. 2018). The reason that recovery of financial costs is not that important is the high maintenance and operation costs of desalination plants. The costs associated with the utilization of desalination plants are partially offset through recuperation from other services, as is also the case on the island of Lampedusa (Palmeros Parada et al. 2023).

Table 5 and Fig. 4 show that financial cost recovery is much more important than non-financial cost recovery (environmental and resource costs). This is because they are calculated with economic information provided by state administration. Non-financial recovery is estimated through indirect methods and, in many cases, it is difficult to reach a numeric value.

Fig. 4
figure 4

Source: second cycle of the Canary Islands River Basin Management Plans (Results are expressed in million Euros)

Representation of the costs calculated for each of the Canary Islands.

Full size image

Figure 4 shows that resource costs represent a very small portion of costs; they are difficult to compute, as they are estimated as the cost of scarcity. They have only been considered in Gran Canaria, La Palma and El Hierro. Gran Canaria presents the highest resource cost recovery among the islands, representing 98% of the resource costs calculated. Environmental cost recovery has been estimated from the measures taken to improve water services according to Article 2.38 of the WFD. In general, as seen in Table 5, it is minor compared to financial cost recovery. It can be deduced from the data that there is an investment to improve water services but not enough economic recovery from the instruments cited in Table 2.

Although calculating environmental costs may be challenging, it is necessary, especially in territories such as islands, which are more vulnerable to climate change (Hernandez et al. 2018) and to consider its effects on groundwater. This is due to the deterioration caused by the intrusion of seawater, a result of rising sea levels, decreased precipitation, and other projections that need to be included and that affect numerous island territories (Gohar et al. 2019).

In Table 5, it can be seen that Tenerife island recovers more environmental cost, meaning that the investments the island made to improve infrastructures for water services are compensated with public and private rates. For its part, El Hierro island does not perceive any recovery from environmental costs (see Table 5) but it is expected that, for the third cycle of the River Basin Management Plan, the proposed measures (improving water infrastructure, diagnosing cost recovery by the services, controlling surface and groundwater extraction or improving wastewater management) will improve these results in the future (CIAEH 2018). It is expected that the next planning cycle (2022–2027) will show improved environmental costs, thanks to implementation of the measures proposed in the second cycle that are mainly based on reduction of losses in water networks and improvement in production and maintenance at desalination plants.

Cost recovery fluctuates significantly across islands and among different water services. Table 6 indicates the absence of available information on cost recovery beyond public networks, except for Tenerife island. This is due to the difficulty of obtaining information from private entities and individuals. As for self-service, it is difficult to estimate, as in most cases it is performed by private entities that do not provide information.

Table 6 Cost recovery for each island and type of water service
Full size table

In general, upstream surface water services recover a small amount of money because of low surface water availability due to irregular distribution of precipitation (Sánchez-Benítez et al.. 2017) and the elevated costs of creating infrastructure that will not provide much water and will not be amortized. In the case of Gran Canaria island, cost recovery for urban water supply is elevated. This is a representation of the efficiency in supply services and the efficient application of tariffs to users.

Desalination costs as a differentiated service were difficult to estimate, due to the low disaggregated information available, since most of the installed desalination capacity is managed by supply service concession companies or by municipalities that also directly provide the service (CIALG 2018). For this reason, part of the desalination costs has been included in the urban water supply. Upstream water services show a lower recovery (see Table 6), as they are mainly controlled by private entities, and the information was not available.

Fuerteventura island does not have any cost recovery related to upstream water services, as there is little information related to how exploitation of the resource is made (CIAFV 2018). The most important cost recovery comes from desalination (see Table 6), although it also accounts for most of the operation and maintenance costs. There is a strong need, in the case of this island, to improve control of losses in the water network system and to improve the mechanisms that control water production and supply. In the case of Lanzarote island, water supply comes from desalination, and costs for production and distribution are elevated, especially regarding energy consumption in the production process (CIALZ 2018). Table 6 shows that water reuse also represents a quite significant amount of cost recovery and is mainly used for agricultural purposes. There are no large volumes without invoicing on the island (CIALZ 2018).

Tenerife’s River Basin Management Plan contains a very detailed analysis of cost recovery from all the water services and sources of information. Table 6 shows that highest cost recovery is due to groundwater upstream services, and urban water supply as groundwater infrastructure is well established (Custodio et al. 2016), meaning that less inversion is required in the water network. There is also an important cost recovery from desalination, as only the costs for operation and maintenance of the plant have been considered. The low-cost recovery for water reuse in Table 6 is due to investments that had been made to improve the water network infrastructure, in order to transport water from urban areas, where it is produced, to the agricultural fields, where it is mainly used (CIATF 2018).

On the island of La Palma, cost recovery for the services of water use, in general, are not very high; they represent about 47% for urban water supply and 51% for agricultural uses. There is a strong need, in this case, for impulse measures that improve cost recovery mechanisms. For La Gomera island, cost recovery is very low also, from downstream irrigation water distribution gathering in 1.85 M€. It constitutes a special case in terms of water management since, in general, it is not charged for most uses, which means that cost recovery is low (CIALG 2018). In this case, it is very important to at least improve financial cost recovery and establish a better system, in which the water consumed in both urban and agricultural services is billed.

In El Hierro island, the major cost recoveries are perceived to be from desalination and urban water supply. Cost recovery is higher for desalination than for upstream services, despite the fact that operation and maintenance costs can be due to the difference between water produced/bought and the water that is invoiced (CIAEH 2018). In the case of this island, it would be advisable to establish not only measures for recovery of financial costs but also for recovery of environmental costs.

Regarding cost recovery from collection and treatment in public networks, Tenerife island shows the highest recuperation, while El Hierro island has the lowest (see Table 6). The volume of treated wastewater has increased considerably since the 1990s, due to the continuous growth of water consumption and improvements in the regulation of wastewater treatment, regardless of its use (Khemlani et al. 2021). In most cases, water tariffs for the users do not contemplate how polluted the water is, household characteristics or volume of treated water (Hoque and Wichelns 2013; Molinos-Senante et al. 2013; Pinto and Marques 2015). Table 6 shows that, at least for El Hierro, La Gomera, La Palma, Fuerteventura and Lanzarote islands, it is urgent to improve cost recovery from wastewater systems by reviewing the tariffs and fees to the users.

Even though water reuse is increasing in the Canary Islands due to high rates of population and tourism, as well as improvement in technologies (Ruiz-Rosa et al. 2020), in general, cost recovery from this service is still low. It would be expected to grow in the future, not only because of higher water demand but also to accomplish the principles of the WFD. The limited integration of treated water into the water cycle is a common issue experienced in other islands around the world, such as in the cases of Crete (Tzanakakis et al. 2022), Taiwan (Shiu et al. 2017), or Mayotte (Mégevand et al. 2021).

Finally, it is difficult to compare upstream and downstream water supply for all the islands, as not all of them share the same system. Although the same methodology has been followed for each island, the information available and the cost recovery mechanisms implemented by water agents differ.

4 Conclusions

Cost recovery is a very useful tool that contributes to reaching good status of water bodies by the ‘Polluter Pays’ principle. The methodology for its establishment is based on three main steps: estimation of financial costs, estimation of environmental costs and estimation of resource costs once the water uses and services are described. Nevertheless, the estimation of each cost is complicated, due to many sources of information, and it depends on ease of access. In most cases, this data is not up to date or it is privately owned. Also, the insular character of the study area makes it difficult to compare, in general, cost recovery in all of the islands. Water management is different because water availability, sources and production mechanisms are not the same in all of them.

The divergent results obtained following the same methodology are due to the different public prices or rates on each island. For instance, La Gomera and El Hierro should review this policy of water prices, as there are almost no charges to the consumers for use of the water. The islands with higher populations (Gran Canaria and Tenerife) showed the best cost recovery results, since they have better and more mechanisms to charge for water services. Lanzarote and Fuerteventura showed the best cost recovery from desalination, as it is their main source of water supply, and it is well established. It is also crucial to improve cost recovery from wastewater collection and treatment, in order to have a more readily available volume of reused water.

Finally, it is noteworthy that our study introduces the novel application of the stochastic frontier model, which allows for the assessment of the efficiency of various water production techniques. This is particularly relevant in the Canary Islands' insular context, where diverse sources such as desalination and groundwater are prevalent. It is imperative to continue conducting studies like these to ensure the proper implementation of the Water Framework Directive in island spaces and to enhance cost recovery.