Privatising Water Utilities and User Perception of Tap Water Quality ...

Water Resour Manage DOI 10.1007/s11269-015-1164-y

Privatising Water Utilities and User Perception of Tap Water Quality: Evidence from Spanish Urban Water Services Miguel A. García-Rubio 1 & Cecilia Tortajada 2 & Francisco González-Gómez 1

Received: 4 July 2015 / Accepted: 12 October 2015 # Springer Science+Business Media Dordrecht 2015

Abstract This paper seeks to contribute to the current debate about public versus private management of urban water supply services. The main purpose of urban water supply is to ensure the provision of a sufficient quantity of good quality water to users. This article examines water quality by using a subjective indicator: user satisfaction with tap water quality. The goal of the paper is to determine whether users perceive a difference in tap water quality supplied by public operators on the one hand, and by private-sector operators on the other. Data is sourced from a survey carried out in 64 Spanish cities. A number of different subjective determinants of water quality have been considered: the socio-economic characteristics of those surveyed, objective indicators of water quality, as well as environmental and economic factors of urban water services. Data analysis, using an ordered logit regression model, shows that when urban water services are in the hands of a private company, the quality of the tap water, as perceived by users, deteriorates. Keywords Quality of tap water . Subjective perception . Urban water services . Privatisation . Ordered logit model . Spain

* Miguel A. García-Rubio [email protected] Cecilia Tortajada [email protected] Francisco González-Gómez [email protected] 1

Department of Applied Economics, Institute of Water Research, University of Granada, Spain, Faculty of Economics and Business Administration, Cartuja Campus, 18071 Granada, Spain

2

Institute of Water Policy, 469C Bukit Timah Road, Oei Tiong Ham Building, Singapore 259772, Singapore

García-Rubio M.A. et al.

1 Introduction The wave of privatisations that began in the late 1970s and early 1980s also went on to include certain local public services such as the collection and treatment of solid urban waste, urban transport and water utilities. Among the local public services, the privatisation of the urban water supply is the one that has been most controversial. This could be due to the fact that the supply of water is considered a merit good, or the difficulty of introducing competition into an industry that is structured around local monopolies (Littlechild 1988). In fact, for many countries this is not even an option (OECD 2004). Within Europe, however, England and Wales, France, Spain, Italy and Portugal, have all set up legal frameworks that open the door to private-sector involvement in the urban water industry. The traditional argument is that private management is more efficient than its public counterpart; nevertheless, over the years there has been no conclusive proof of this.1 Applied research suggests that ownership (private or public) of the operator is irrelevant when it comes to promoting efficiency, while the regulatory and institutional framework in which it operates, on the other hand, is a determining factor (GonzálezGómez and García-Rubio 2008; De Witte and Marques 2010). In this case, there are examples of both successful and failed cases of privatisation (Kanakoudis and Tsitsifli 2014). There is currently a certain amount of opposition to the privatisation of the urban water industry, as can be seen in the remunicipalisation that has taken place in cities such as Paris and Berlin (Hall et al. 2013), the introduction of national regulations in certain countries where such privatisation is prohibited –The Netherlands, Uruguay and Ecuador– (Marques 2010), or citizen resistance in response to announcements of privatisation, such as in Italy and Bolivia (Lobina et al. 2011; Fattori 2013). Detractors of private-sector involvement in the urban water industry argue that the main goal of a private company is to make a profit rather than to serve the interests of the general public, and this could imply higher rates and lower quality services (Lobina 2005). This could also result in lower incentives to pursue water conservation(Barrett and Wallace 2011; Romano et al. 2014). There is empirical evidence to suggest that the price of drinking water is higher under private-sector management in France (Chong et al. 2006; Carpentier et al. 2007), Spain (Martínez-Espiñeira et al. 2009), Germany (Ruester and Zschille 2010) and Italy (Guerrini et al. 2011). Nevertheless, in the case of Spain, and excluding direct control by the municipalities, the results of García-Valiñas et al. (2013) also indicate that the price of water is higher with public-sector operators; and in Italy, Romano et al. (2015) find no significant difference in price according to ownership. The market structure in the private sector of the industry also appears to be a determining factor in price differences (Bel et al. 2015). When analysing the impact of ownership on the efficient management of urban water services, the quality of the service has rarely been taken into consideration.2 In the few cases that do exist, a number of research papers conclude that there is a trade-off between efficiency and quality of the services. In some cases, private companies are more efficient, but to the detriment of the state of repair of the networks (García-Rubio et al. 2010); in others, researchers have found that privatesector involvement leads to a fall in investments levels (Bakker 2010; Romano et al. 2013). Both results could be detrimental to the quality of the services provided. 1

Recent research that analyses the relationship between management ownership and efficiency of service provision in urban areas include González-Gómez et al. (2013) and Guerrini et al. (2015). 2 Nevertheless, there are case studies that conclude that privatisation leads to a lower quality service, among other things (Pigeon et al. 2012). And Bontemps and Naugues (2009) find no differences in users' decision on whether or not to consume tap water relating to ownership of the operator.

Private Water Utilities and User Perception of Tap Water Quality

There has been still less research analysing the relation between ownership and the quality of the water supplied and in this case, the evidence is not conclusive. Destandau and Garcia (2014) find that even when public management initially incurs higher operational and maintenance costs –which include water treatment costs–, the findings are refuted when the concentration of nitrates in the water supplied is used as a quality indicator. Nevertheless, Wallsten and Kosec (2008) do not find that ownership is a determining factor when explaining differences in non-compliance with regulations governing the quality of drinking water. Dore et al. (2004) however indicate that following privatisation of the sector in the United Kingdom and Wales, the percentage of water analysis that complied with regulations increased; according to the competent British authority, the Drinking Water Inspectorate (see http://dwi.defra.gov.uk/), the rise in the number of analyses meeting requirements was due to a rise in the number of checks being carried out. An adequate monitoring and control policy might dissuade non-compliance with the regulations that apply to drinking water quality standards, regardless of the type of management. This might potentially explain, in certain cases, why it is more difficult to find differences in the quality of the water based on ownership. Starting from the premise that the main aim of urban water supply is to ensure provision of a sufficient quantity of good quality drinking water, this research paper seeks to contribute to the current debate regarding potential differences in the quality of water supplied according to the type of management –public or private. It also aims to confirm whether users perceive a lower quality of water when the company supplying the water is from the private sector. The quality of water is not, therefore, analysed using objective indicators but rather the dependent variable is subjective satisfaction with tap water quality. With this objective, we must rule out the influence of socioeconomic and ideological characteristics of users, the objective indicators of tap water quality that cannot be controlled by the operator, as well as the environmental factors that might affect the quality of water at its point of origin. To the best of our knowledge, this is the first study that analyses the relation between ownership of the service provider and the quality of tap water introducing an indicator of subjective satisfaction as a dependent variable. Information is sourced from 1023 household surveys from 64 Spanish cities with populations of over 100,000; data were obtained by way of a survey carried out by the Centre for Sociological Research (CIS) in 2011. The ordered logit regression technique is used to carry out the analysis. The results show that when the management of urban water services is in the hands of private companies, users express lower levels of satisfaction with regard to tap water quality. The rest of the paper is structured as follows. The second section describes the institutional framework of the privatisation of urban water services and the quality control of drinking water in Spain. The data used and the methodology are described in section three. Section four contains the results of the analysis and discusses the results. The paper concludes with the main findings of the study.

2 Institutional and Legislative Framework 2.1 The Privatisation of Urban Water Management in Spain Local Government Regulatory Law (Ley de Bases del Régimen Local) 7/1985, of 2nd April, establishes that municipalities are responsible for urban water services. Article 85 of the law states that local public services should be managed in the most sustainable and efficient manner.With this purpose, in addition to direct control –either by the local authority itself or through a public


corporation- it establishes that the municipality can privatise service provision following what is established in the Royal Decree 3/2011, of 14th November, and the consolidated text of the Public Sector Contracts Act, −through a concession, public-private partnership, lease contract and management contract. Currently, private companies manage urban water services in 23 % of Spanish municipalities (González-Gómez et al. 2014), although, in terms of population served, the participation of private companies in the industry stands at 55 %. Private-sector management therefore has a relatively greater presence in the more populated municipalities. Local authorities tend to opt for privatisation in more populated municipalities where management of the service is more complex; for their part, privates companies are more interested in these municipalities, as expected profits are higher (González-Gómez et al. 2011). Consequently, this research paper focuses on cities with a population of over 100,000. The water industry is the industry that probably comes closest to a monopoly situation. Similar to the French model, in Spain, competition only exists between companies when local authorities put the management of water supply services out to tender. Once the tender process has concluded, the company awarded the service contract behaves like a local monopoly throughout the term of the contract. Under any of the accepted management formats, the privatisation of water supply services is for a specified length of time. This lack of competition is compounded in Spain for two reasons (González-Gómez et al. 2014). Firstly, in the Spanish urban water services industry, there are two clearly dominant companies: Aqualia and Aguas de Barcelona. Secondly, there is a high spatial concentration of certain companies in certain areas of the country: for example, Aguas de Barcelona in Catalonia, Aguas de Valencia in the province of Valencia and FACSA in the province of Castellon. One way to compensate for the possible effects of lack of competition in the sector is to create regulatory bodies. Nevertheless, in Spain there is still no national regulatory body in place. Consequently, in municipalities where water management has been delegated to private companies, the only monitoring of activities are those carried out by the local authority itself. As a result, in the case of Spain, the impacts of lack of competition might be exacerbated by information asymmetries that work to the advantage of the private companies, or by the capture of the local regulatory authority. The potential weakness of the effective control of private management makes for a perfect platform from which private companies can focus exclusively on maximising profits, rather than acting in the interests of the general public. Thus, one possible effect is that private companies provide a lower quality service given that they are strongly motivated to reduce costs, but have few incentives to improve the quality of the services they provide (Destandau and Garcia 2014).

2.2 Quality Control Of Water For Human Consumption In Spain The main goal of an urban water service is to ensure the provision of a sufficient quantity of good quality water to users. In Spain, urban water services have to follow quality standards stipulated in Royal Decree 140/2003, of 7th February, which establishes the sanitary criteria of the quality of drinking water. It describes the substances that can be used to treat water for human consumption, the minimum and maximum parametric values of substances that must be monitored during water analysis microbiological parameters, chemical parameters, indicator parameters and radioactivity– the sampling points and the frequency of analysis. Controls should be carried out as water leaves the drinking water treatment plant or the header tank, when leaving the regulating and/or distribution tanks, and in the distribution network. At these


sampling points, the minimum frequency of controls depends on the volume of water treated or distributed. Water should also be checked at the consumer’s tap; here, the minimum frequency of controls depends on the number of inhabitants receiving the supply. The parameters that require monitoring at the consumer’s tap are the organoleptic characteristics (smell, colour, taste and clarity), conductivity, pH, ammonium, combined and free residual chlorine, microbiological parameters (faecal coliforms), and specific metals when there are suspicions that the installations contain this type of material (copper, chromium, nickel, iron and lead). Water quality monitoring is the responsibility of those running the urban water services; the municipalities have to ensure that the water supplied is suitable for human consumption. With the exception of a few, specific cases, which are detected during the monitoring activities of the National Information System of Water Consumption (SINAC), water for human consumption in large urban water supplies in Spain complies with the required legal standards. As already mentioned, the Spanish institutional framework provides few incentives to improve the quality of services; nevertheless, it would seem logical that private companies do not reduce spending in budgetary line items as doing so may lead to failure to comply with legal standards. On the one hand, any infringement would mean incurring fines; on the other, it should be taken into account that any fall in quality could be a reason for there municipalisation of the service. Nevertheless, meeting legal standards does not necessarily imply that the quality of water supplied to the households of different cities is exactly the same. Although still suitable for human consumption, there may be organoleptic differences in the water supplied by different management units. These differences, although in compliance with the parameters stipulated in the regulations, are the differences in water quality that are perceived by users.3 The organoleptic differences in tap water quality may vary from one urban water service to the next for reasons that come under the responsibility of the operator. On the one hand, there may be differences in the treatment of raw water at the water treatment plant. Furthermore, the state of repair of the water supply network also affects the quality of the water supplied. If the network is in a bad state of repair, with leaks, cracks and frequent interruptions in the water supply, there will be high volumes of water loss; but in addition, the lack of water-tightness of the pipes –and the frequent drop in pressure in response to such cases so as to reduce water losses–, can also result in lower water quality and in different organoleptic characteristics by the time it reaches the consumer’s tap. Therefore, the hypothesis that we are going to contrast in this research paper is the following: Ho: Levels of user satisfaction with tap water quality are lower when the service is managed by a private company.

3 Data and Methodology In order to carry out research on the determining factors of the user perception of tap water quality, we used the results of a questionnaire-based survey. The survey was carried out between 29th June and 19th July 2011 by the Centre for Sociological Research (CIS 2011), the Obviously there are other more important parameters from the perspective of the sanitary quality of water – heavy metals, volatile organic compounds, trihalomethanes, pesticides, cyanotoxins, etc. At low concentrations, however, their presence is difficult for users to identify directly; it is therefore very unlikely that these influenceone way or another users' perception of tap water quality (Turgeon et al. 2004).

3


aim of which was to analyse the opinions of citizens on the quality of public water supply services in Spain. The total of the sample is represented by the total population (both males and females above 18 years old). The questionnaires were filled out during personal interviews carried out in the homes of the respondents. It’s applied a process of multistage sampling, stratified by conglomerates, with selection of proportional random sampling of primary units (municipalities), secondary units (sections), and last units (individuals) by random routes and sex and age quotes. The results show the level of satisfaction on tap water quality as well as the socio-economic and ideological characteristics of the respondents. Given the ability to identify the geographical location of the respondents, we also know which operator provides each respondent’s with urban water services. Data from secondary sources have been used to fill out any details provided during the survey, including objective indicators as to the quality of the water, as well as economic and environmental factors of the corresponding urban water services. User satisfaction with tap water quality (SATISQUAL) is obtained using the question BAre you satisfied with your tap water quality?^ The respondents scored their perception of tap water quality on a scale from 1 to 4, from Bvery unsatisfactory^ to Bvery satisfactory^. The ordered logit model is suitable for this type of dependent variable; said model is commonly presented as a latent variable model. Correlation between latent variable, yi*, and the independent model variables, xi, is: y*i ¼ xi β þ εi

ð1Þ

where ε is a random error. The response categories are linked to the latent variable by the following measurement model: yi yi yi yi

¼ 1 ðvery unsatisfactoryÞ if τ 0 ¼ −∞≤y*i < τ 1 ¼ 2 ðunsatisfactoryÞ if τ 1 ≤ y*i < τ 2 ¼ 3 ðsatisfactoryÞ if τ 2 ≤ y*i < τ 3 ¼ 4 ðvery satisfactoryÞ if τ 3 ≤ y*i < τ 4 ¼ ∞

The standard formula for predicted probability in the ordered logit model is: Prðy ¼ mjxÞ ¼ Pr τ m−1 < y*i ≤τ m ¼ F ðτ m −xβ Þ− F ðτ m−1 −xβ Þ

ð2Þ

ð3Þ

where F is the cumulative distribution for ε, and logistically distributed –with a mean of 0 and a variance of π2/3. The marginal change in probability is calculated as: ∂Prðy ¼ mjxÞ ∂ F ðτ m −xβÞ ∂F ðτ m−1 −xβ Þ ¼ − ∂xk ∂xk ∂xk

ð4Þ

Model estimation is carried out with maximum likelihood using interactive optimisation algorithms. A general overview of the ordered logit model, from a theoretical perspective, can be found in Wooldridge (2002). For the applied perspective, see Scott and Freese (2014). The number of cases in the sample –those interviewed– is 1023. Information is available for 64 Spanish cities, including the 82.5 % of Spanish municipalities with a population of over 100,000.The population of the cities in the sample is between 107,597 and 3,233,527 inhabitants, and includes the two most populated cities in the country: Madrid y Barcelona. They are also representative of the different environments where water utilities operate in Spain (water sources, geographical location, precipitation, average temperature, etc.). In 53.8 % of the municipalities, the urban water services are provided by a private company, while ownership of the rest is in the hands of the public sector. Table 1 includes the descriptive statistics of the sample, and defines the variables used in the analysis.

Private Water Utilities and User Perception of Tap Water Quality Table 1 Descriptive statistics of sample Variable

Description

Mean

S.D.

Min

Max

SATISQUAL

Respondent’s satisfaction with tap water quality on a scale 1 to 4 (1 = very unsatisfactory; 4 = very satisfactory)

2.759

0.802

1

4

GENDER

1 if respondent is female, 0 otherwise

0.518

0.499

0

1

AGE

Respondent’s age in years

47.136

17.664

18

91

EDULEVEL

Respondent’s educational level (1, higher education; 0, otherwise)

0.349

0.477

0

1

FAMINCOM

Respondent’s monthly household income (1, high income –over €4500/month–; 0 otherwise)

0.028

0.166

0

1

SATISPPSS

Respondent’s satisfaction about all public services –provided by State, Autonomous Regions or local authorities– (1, very satisfactory; 0, otherwise)

0.029

0.169

0

1

POLIDEOL

Respondent’s political ideology (1, left-wing ideology; 0, otherwise)

0.169

0.375

0

1

HARDNESS

Raw water hardness in the locality (°F)

19.044

13.808

0.6

60

RESICHLO

Free residual chlorine in the user’s tap (mgr/l)

0.527

0.304

0.05

1.2

COASTAL

1 if municipality is coastal, 0 otherwise

0.431

0.495

0

1

Average annual rainfall (l/m2) Company ownership providing urban water service (1, private; 0, public)

670.985 0.538

275.823 0.499

279.93 0

1511.85 1

Water price for a consumption of 10.8 m3/household/month (€, excluding taxes)

9.885

5.046

3.23

32.75

RAINFALL OWNERSHIP

WPRICE

The questionnaire used also includes other characteristics of the respondents, such as gender (GENDER) –51.8 % of respondents are women–, age (AGE), education level (EDULEVEL), household income (FAMINCOM), level of satisfaction with all public services (SATISPPSS) and political leanings (POLIDEOL).4 As per Turgeon et al. (2004), the respondents were classified into two groups according to their level of education (EDULEVEL): higher education and others; 34.9 % of respondents have a higher education qualification. With regard to levels of household income (FAMINCOM), high-income families are expected to be those most likely to adopt expensive measures to mitigate the bad quality of tap water; 2.8 % of respondents reported they had a household income of more than €4500 per month. The level of satisfaction for all public services is represented by SATISPPSS; some 2.9 % of respondents consider public services as a whole to be Bvery satisfactory^. The higher the level of satisfaction with public services in general makes it more likely that respondents will also register higher levels of satisfaction with tap water quality. Lastly, POLIDEOL considers the political leanings of respondents. Left-wing respondents are more likely to reject the private management of urban water services. Water hardness at its point of origin in the municipality –measured in French degrees of hardness– (HARDNESS) and the average annual level of residual chlorine a the consumer’s water tap –in mg/l– (RESICHLO) are included as objective indicators of water quality.5 Data are sourced from 4

The questionnaire also provides information about their employment status, marital status, nationality and religion. However, none of these variables were significant in our analysis. 5 All samples are free of microbial contamination –faecal coliforms. Furthermore, the organoleptic characteristics of the water could not be included in the analysis given that figures were only available for sporadic analysis, and said parameters can change over time even within the same urban water service.


tap water quality control analysis carried out by the service operators, taken from the information they provide on their websites. However, the first variable is not under the control of the urban water service provider; in Spain, water Bsoftening^ treatments at the drinking water treatment plants are not very frequent. Both variables are frequently included in papers that study users' perception of tap water quality (Turgeon et al. 2004). Lastly under consideration are the economic variables of urban water services and the variables that relate to the environment where they operate. The economic variables considered are ownership of the companies that manage the urban water services –public or private– (OWNERSHIP) and the price of water for a household with a monthly consumption level of 10.8 m36 (WPRICE). To be able to analyse if the type of ownership of the company providing urban water services influences the user’s perception of tap water quality, the characteristics of the environment where said companies operate must be taken into consideration. Otherwise, operators in unfavourable environments would be penalised (e.g., polluted or scarce raw water) (González-Gómez and García-Rubio 2008; PicazoTadeo et al. 2009). In this case, the characteristic of the environment that might affect the quality of the water at its point of origin, and which is beyond the control of the companies providing the urban water services must be taken into consideration. Consequently, this study considers whether the municipality is on the coast or not (COASTAL) and also the average annual rainfall of the municipality in l/m2 (RAINFALL).7 All economic and environment variables were sourced from the companies that operate the urban water services, with the exception of the figures for average annual rainfall, which were obtained from the Spanish National Meteorological Agency (AEMET 2015).

4 Results and Discussion The analysis of the impact that the four types of determinants considered –socio-economic and ideological characteristics of respondents, objective indicators of water quality, and the economic and environment factors of urban water services– have on users' perception of tap water quality produces some interesting results. The results of the four estimated models can be seen in Table 2. Model 1 only includes the socio-economic and ideological characteristics and the objective indicators of water quality. The GENDER, AGE and POLIDEOL variables are not statistically significant; Dogaru et al. (2009) shows similar results with the relation to the age and gender of respondents. In the broader risk perception literature, the role played by age remains ambiguous and is likely to be hazard-dependent (Doria 2010). With regard to the variables in the analysis that were statistically significant, EDULEVEL has a negative impact; respondents with a further education qualification find tap water quality to be less satisfactory. There may be a direct relation between the education level and perception of chemical contamination of drinking water (Auslander and Langlois 1993) whereby these users would be more aware of the risks associated with poor quality water, in turn making them more demanding. With regard to FAMINCON, respondents with higher levels of household income express a greater level of satisfaction with tap water quality. Households within higher income levels have a 6

This figure is the average water consumption per household in Spain, based on the average per capita consumption −142 l/inhabitant/day– and the average household size −2.53 inhabitants/household– (INE 2013). Other variables are considered, such as the number of users of the service and whether or not the provision of urban water services comes under an association of municipalities; these variables could be considered proxies of the economies of scale in the sector (González-Gómez and García-Rubio 2008). The tourism index of the municipality is also considered (Caixa 2011). However, none of these variables were statistically significant in the analysis. 7

z-statistic in brackets

***Significant at 1 %

**Significant at 5 %

*Significant at 10 %

Interaction term POLIDEOLxOWNERSHIP Observations Log likelihood Pseudo R−squared LR Chi-squared (p-value)

WPRICE

OWNERSHIP

Economic factors

RAINFALL

COASTAL

Environmental factors

RESICHLO

HARDNESS

Water quality

POLIDEOL

SATISPPSS

FAMINCOM

EDULEVEL

AGE

GENDER

Respondent’s characteristics

Variable

Table 2 Estimated ordered logit models

1023 −1042.676 0.1081 252.76 (0.000)

−0.0551*** (−11.48) −1.7440*** (−8.00)

−0.0904 (−0.73) 0.0012 (0.33) −0.2939** (−2.19) 0.9026** (2.40) 1.6468*** (4.29) 0.0370 (0.22)

Model 1

1023 −999.096 0.1454 339.92 (0.000)

−1.3147*** (−8.25) 0.0019*** (7.00)

−0.0318*** (−5.78) −1.5198*** (−6.44)

−0.1202 (−0.96) 0.0014 (0.40) −0.3251** (−2.39) 0.7799** (2.07) 1.6453*** (4.29) 0.1806 (1.07)

Model 2

1023 −987.357 0.1554 362.40 (0.000)

−0.6780*** (−3.92) −0.0596*** (−3.93)

−1.0450*** (−6.09) 0.0012*** (3.82)

−0.0326*** (−5.42) −0.9539*** (−3.59)

−0.1145 (−0.91) 0.0015 (0.42) −0.3610*** (−2.64) 0.7764** (2.04) 1.5931*** (4.11) 0.1631 (0.96)

Model 3

0.0478 (0.14) 1023 −987.347 0.1554 363.42 (0.000)

−0.6707*** (−3.72) −0.0596*** (−3.93)

−1.0443*** (−6.07) 0.0012*** (3.80)

−0.0326*** (−5.43) −0.9539*** (−3.59)

−0.1150 (−0.91) 0.0015 (0.42) −0.3623** (−2.65) 0.7765** (2.04) 1.5941*** (4.11) 0.1419 (0.63)

Model 4



greater inclination to implement expensive measures to improve drinking water quality (Larson and Gnedenko 1999), such as installing filters and reverse osmosis systems (Gartin et al. 2010) or regularly purchasing bottled water for drinking and cooking (Yoo 2005). These households therefore have safe alternatives for the uses of water and are likely to be more demanding as far as quality is concerned, which makes it more likely that they have a better understanding of the importance of tap water quality. Lastly, the respondents who consider public services in general to be very satisfactory (SATISPPSS), show a greater level of satisfaction with tap water quality; this can most likely be put down to ideological reasons. With regard to the objective indicators of water quality, when HARDNESS increases, the degree of satisfaction of water quality falls. Although the hardness of drinking water does not appear to have adverse effects on health, it is important for consumers' acceptability and for economic and operational reasons (WHO 2011). In fact, in Spanish regions where hard water is more prevalent, there are higher levels of consumption of bottled water per capita (MAGRAMA 2011). Furthermore, the higher the RESICHLO, the lower the satisfaction levels of tap water quality; this negative effect of the level of free residual chlorine on the subjective perception of water quality is a frequent finding in the literature (Doria 2010; Piriou et al. 2014). Model 2 keeps the same variables as those from Model 1 but it also includes the environment variables that are beyond the control of urban water service providers. COASTAL has a negative and statistically significant impact on subjective satisfaction of tap water quality. In 70 % of Spanish urban water services, the water comes from groundwater sources, and in the coastal waters, nitrates and chlorides are frequently present (López-Geta and López-Vera 2006). The conductivity of water at its point of origin –in μS/m– in the 64 municipalities included in the study, can be found in the drinking water quality analyses carried out by the urban water services.The results of a Mann-Whitney test made (Z = −3.079, pvalue = 0.002) show that there are statistically significant differences in conductivity between coastal and inland municipalities from the sample. These differences in the salinity of the water may affect the perceived quality of tap water. With regard to RAINFALL, respondents living in regions with higher rainfall are more satisfied with tap water quality. For a same-depth extraction of raw water from the natural environment, the pressure on water resources will be less for regions with a greater rainfall.8 The unregulated use of water resources could lead to a fall in the quality of water, especially in the case of underground water. Lastly, Model 3 also includes the economic factors in the analysis. The price of water (WPRICE) is significant and inversely related to the level of satisfaction of tap water quality; this suggests that the respondents take the price of water into consideration when responding to questions about their satisfaction of tap water quality. Existing literature indicates that the price charged for urban water services might be higher when operated by private companies (Chong et al. 2006; Carpentier et al. 2007; Martínez-Espiñeira et al. 2009; Ruester and Zschille 2010; Guerrini et al. 2011); although according to García et al. (2005)these results require further qualification given that price level might be linked to operator strategy, not merely the type of ownership. In the urban water services of our sample, the average price for the consumption of 10.8 m3/household/month is €10.06/m3 and €9.48/m3 for public and private-sector operators, respectively; the results of a Mann-Whitney test (Z = 0.585, p-value = 0.558) show that there

8

Furthermore, in Spain, the regions with heaviest rainfall have lower average annual temperatures. Therefore, the per capita extraction of water is less than that of drier regions.


are no statistically significant differences in the price of water charged by the public and private-sector operators in our sample. With regard to OWNERSHIP, when the management of urban water services is in the hands of a private company, the subjective perception of tap water quality is worse. This finding might be due to ideological reasons; in Doria et al. (2009), tap water use is explained by trust in water companies. Respondents' negative response to the private management of urban water services might have an influence over their stated perception of tap water quality –protest responses. This type of behaviour would most likely be associated with respondents with left-wing political leanings. To analyse whether this potential effect is real, Model 4 also includes the POLIDEOLxOWNERSHIP interaction term; results show that said term is not statistically significant (Table 2). Therefore, the inverse relation and statistical significance between the perception of water quality and private-sector management do not appear to be for ideological reasons. The marginal probabilities for Model 3 appear in Table 3. With regard to the OWNERSHIP variable, the privatisation of the management of urban water services is significantly and negatively (or positively) related to the probability of considering tap water quality to be satisfactory or very satisfactory (or unsatisfactory or very unsatisfactory). More specifically, when the management of urban water services shifts from the public to the private sector, the probability of users finding tap water quality satisfactory or very satisfactory falls by 6.71 % and 5.66 %, respectively; alternatively, the probability of users finding tap water quality unsatisfactory or very unsatisfactory rises by 9.39 % and 2.99 %, respectively. As already stated, this result does not appear to be related to ideological factors of Table 3 Marginal effects in percentages for Model 3 Variable

Very un-satisfactory

Un-satisfactory

Satisfactory

Very satisfactory

Respondent’s characteristics GENDER

0.51 (0.91)

1.61 (0.91)

−1.19 (0.91)

−0.93 (−0.90)

AGE

−0.01 (−0.42)

−0.02 (−0.42)

0.02 (0.42)

0.01 (0.42)

1

1

1.69** (2.45)

5.17** (2.57)

−4.06** (−2.41)

−2.80*** (−2.70)

FAMINCOM

1

−2.52*** (−2.78)

−9.18** (−2.53)

3.22*** (2.59)

8.47 (1.58)

SATISPPSS

1

−3.85*** (−6.16)

−15.10*** (−6.78)

−3.99 (−0.67)

22.91*** (2.80)

POLIDEOL

−0.69 (−1.00)

−2.25 (−0.98)

1.56 (1.05)

1.38 (0.92)

HARDNESS

0.15*** (4.90)

0.46*** (5.20)

−0.34*** (−4.73)

−0.26*** (−5.16)

RESICHLO

4.25*** (3.40)

13.43*** (3.56)

−9.95*** (−3.39)

−7.73*** (−3.54)

EDULEVEL

Water quality

Environmental factors COASTAL

5.13*** (4.93)

14.80*** (5.91)

−11.75*** (−5.19)

−8.17*** (−5.79)

RAINFALL

−0.01*** (−3.63)

−0.02*** (−3.75)

0.01*** (3.57)

0.01*** (3.72)

2.99*** (3.69) 0.27*** (3.74)

9.39*** (3.95) 0.84*** (3.83)

−6.71*** (−3.80) −0.62*** (−3.64)

−5.66*** (−3.74) −0.48*** (−3.82)

1

Economic factors 1

OWNERSHIP WPRICE 1

Marginal effect for discrete change of dummy variable from 0 to 1

*Significant at 10 %, **Significant at 5 % ***Significant at 1 % z-statistic in brackets


the respondents. Consequently, considering the environment factors that are beyond the control of the operator, the differences in perceived quality might be related to the better or worst managementpractices of the private company.

5 Conclusions The aim of this research paper is to analyse if users find tap water quality less satisfactory when urban water services are managed by a private company. To examine the subjective level of satisfaction with the quality of water, variables have been used that reflect the socioeconomic and ideological characteristics of users, objective indicators of tap water quality, economic factors of urban water supply and environment factors that are beyond the control of the operator and might affect the quality of water at its point of origin. Information is sourced from surveys of 1023 households in 64 Spanish cities. The results of the ordered logit regression models show that the satisfaction levels of users regarding tap water quality falls when the level of education of the users is high, when household income is not high and when the level of users' satisfaction with public services in general is low. Furthermore, the levels of hardness and free residual chlorine in the consumer’s tap are inversely related to user satisfaction. Regarding the environment factors, the subjective perception of users worsens when the municipality is on the coast –which might be linked to the high levels of salinity of the water being supplied–, and improves the higher the rainfall levels –probably due to there being less pressure on water resources in those areas. With regard to the price of the water supplied, for average household consumption in Spain, the higher the price, the lower the user satisfaction. Taking into consideration the aim of this research paper, the most significant result is that the privatisation of the management of urban water services appears to bring about a fall in the level of user satisfaction with tap water quality; furthermore, this result does not appear to be related to the ideological factors of the users. Although water for human consumption in Spanish urban water supply networks does comply with legal requirements, it could occur a less zealous approach of private companies with regard to the quality of the water they supply; more specifically, with certain characteristics that can be easily identified by users –organoleptic characteristics. The pursuit of maximum profits by private companies, contrary to the interests of the general public, might be bolstered by the Spanish framework due to the weakness of effective control of operators of urban water services. The reasonfor these differences in the subjective perception of users with regard to public and private-sector management might stem from the treatment of raw water at the water treatment plant, or even from the state of repair of the water supply network – water-tightness issues with the water networks and drops in pressure as a measure to reduce water losses. The remunicipalisation that has occurred in recent years in certain Spanish municipalities, due to the failure to invest in the network as stipulated in the contracts (insufficient maintenance of the water supply network), points towards the second hypothesis. The results suggest that it might be appropriate for local authorities, making use of their powers to monitor the management of urban water services, to periodically consult with consumers as to their levels of satisfaction with tap water quality. Any substantial fall in the subjective perception of users, despite compliance with statutory standards for the quality of drinking water, would be indicative of a less zealous approach of the operator towards the quality of the water they supply. There could be provisions in the service specifications of the


public tender process for the privatisation of urban water services, making it an obligation on their part to take measures should this occur. Lastly, any additional advances in this line of research would require overcoming certain limitations in the available data. For example, access to sectorised data on the water supply networks would be essential, with data being broken down according to the year the networks were installed and on their state of repair, identifying the respondents according to the sector of the network that supplies their water. This would allow a comparison to see if the differences in users' satisfaction with tap water quality are really related to an inadequate rate of conservation and network renewal maintenance by the private-sector operator. Unfortunately, the lack of transparency of the urban water industry in Spain and the lack of any national regulatory body make it very difficult to get access to this type of information. Acknowledgments The authors gratefully acknowledge the constructive comments and suggestions from the anonymous reviewers. The authors also gratefully acknowledge the financial support from the following Spanish institutions: the Consejería de Economía, Innovación, Ciencia y Empleo from the Government of Andalusia (P11SEJ-7039 and P11-SEJ-7294), and the Spanish Ministry of Economics and Competitiveness of Spain (Project ECO2012‐32189). The authors also acknowledge the assistance granted under the Programa de Fortalecimiento de las Capacidades de I+D+i, jointly funded by the European Regional Development Fund (ERDF) and the Junta de Andalucía (The Regional Government). Finally, we appreciate the information provided by the Agencia Estatal de Meteorología (Ministry of Agriculture, Food and Environment).

References AEMET (2015) Registros climáticos. Agencia Estatal de Meteorología, Madrid. http://www.aemet.es/es/idi/ clima/registros_climaticos. Accessed 10 May 2015. Auslander BA, Langlois PH (1993) Toronto tap water: perception of its quality and use of alternatives. Canadian J Public Heal 84(2):99–102. Bakker K (2010) Privatizing water. Governance Failure and the World’s Urban Water Crisis. Cornell University Press, New York Barrett G, Wallace M (2011) An institutional economics perspective: the impact of water provider privatisation on water conservation in England and Australia. Water Resour Manag 25(5):1325– 1340. Bel G, González-Gómez F, Picazo-Tadeo AJ (2015) Does concentration affect prices in the urban water industry? Environ Plann C (in press) Bontemps C, Naugues C (2009). Carafe ou bouteille? Lerôle de la qualité de l’environnement dans la décision du consommateur. Econ Prévis 188(2):61–79. Caixa (2011) Anuario económico de españa 2011. Caja de Ahorros y Pensiones de Barcelona, Barcelona Carpentier A, Nauges C, Reynaud A, Thomas A (2007) Effets de la délégation sur le prix de l’eau potable en France. Une analyse à partir de la littérature sur les Beffets de traitement^. Econ Prévis 174(3):1–19. Chong E, Huet F, Saussier S, Steiner F (2006) Public-private partnerships and prices: evidence from water distribution in France. Rev Ind Organ 29(1–2):149–169. CIS (2011) Encuesta de calidad de los servicios públicos (VI). Centro de Investigaciones Sociológicas, Madrid. http://www.cis.es/cis/opencm/ES/1_encuestas/estudios/ ver.jsp?estudio = 12904. Accessed 10 April 2015. De Witte K, Marques RC (2010) Designing performance incentives, an international benchmark study in the water sector. Cent Eur J Oper Res 18(2):189–220. Destandau F, Garcia S (2014) Service quality, scale economies and ownership: an econometric analysis of water supply costs. J Regul Econ 46(2):152–182. Dogaru D, Zobrist J, Balteanu D, Popescu C, Sima M, Amini M, Yang H (2009) Community perception of water quality in a mining-affected area: a case study for the certej catchment in the apuseni mountains in Romania. Environ Manag 43(6):1131–1145. Dore MH, Kushner J, Zumer K (2004) Privatization of water in the UK and France—what can we Learn? Util Policy 12(1):41–50.

García-Rubio M.A. et al. Doria MF (2010) Factors influencing public perception of drinking water quality. Water Policy 12(1):1–19. Doria MF, Pidgeon N, Hunter PR (2009) Perceptions of drinking water quality and risk and its effect on behaviour: a cross-national study. Sci Total Environ 407(21):5455–5464. Fattori T (2013) From the water commons movement to the commonification of the public realm. S Atl Quart 112(2):377–387 García S, Guerin-Schneider L, Fauquert G (2005) Analysis of water price determinants in France: cost recovery, competition for the market and operator’s strategy. Water Supply 5(6):173–181. García-Rubio MA, González-Gómez F, Guardiola J (2010) Performance and ownership in the governance of urban water. Munic Eng 163(1):51–58. García-Valiñas MA, González-Gómez F, Picazo-Tadeo AJ (2013) Is the price of water of residential use related to provider ownership? Empirical evidence from Spain. Util Policy 24(1):59–69. Gartin M, Crona B, Wutich A, Westerhoff P (2010) Urban ethnohydrology: cultural knowledge of water quality and water management in a desert city. Ecol Soc 15(4):36. González-Gómez F, García-Rubio MA (2008) Efficiency in the management of urban water supply services. What have we learned after four decades of research? Hacienda Pública Esp 185(2):39–67. González-Gómez F, Picazo-Tadeo AJ, Guardiola J (2011) Why do local governments privatise the provision of water services? Empirical evidence from Spain. Public Adm 89(2):471–492. González-Gómez F, García-Rubio MA, Alcalá-Olid F, Ortega-Díaz MI (2013) Outsourcing and efficiency in the management of rural water services. Water Resour Manag 27(3):731–747. González-Gómez F, García-Rubio MA, González-Martínez J (2014) Beyond the public-private controversy in urban water management in Spain. Util Policy 31:1–9. Guerrini A, Romano G, Campedelli, B (2011) Factors affecting the performance of water utility companies. Int J Public Sector Manage 24(6):543–566. Guerrini A, Romano G, Leardini C, Martini M (2015) Measuring the efficiency of wastewater services through data envelopment analysis. Water Sci Technol 71(12):1845–1851. Hall D, Lobina E, Terhorst P (2013) Re-municipalisation in the early twenty-first century: water in France and energy in Germany. Int Rev Appl Econ 27(2):193–214. INE (2013) Encuesta sobre el suministro y saneamiento del agua 2012. Instituto Nacional de Estadística, Madrid. Available online: http://www.ine.es/jaxi/ menu.do?type = pcaxis&path = %2Ft26%2Fp067%2Fp01&file = inebase. Kanakoudis V, Tsitsifli S (2014) Doing the urban water supply job: from privatization to remunicipalisation and the third pillar of the performance based service contracts. Water Util J 8:31–46. Larson BA, Gnedenko ED (1999) Avoiding health risks from drinking water in Moscow: an empirical analysis. Environ Dev Econ 1999(4):565–581. Littlechild S (1988) Economic regulation of privatised water authorities and some further reflections. Oxf Rev Econ Policy 4:40–68 Lobina E (2005) Problems with private water concessions: a review of experiences and analysis of dynamics. Int J Water Resour Dev 21(1):55–87 Lobina E, Terhorst P, Popov V (2011) Policy networks and social resistance to water privatization in Latin America. P Soc Behav Sci 10:19–25. López-Geta JA, López-Vera F (2006) Estado del conocimiento de las aguas subterráneas en españa. B Geol Minero 117(1):89–114. MAGRAMA (2011) Base de datos de consumo en hogares. Ministerio de Agricultura, Alimentación y Medio Ambiente, Madrid. Available online: http://www.magra-ma.gob.es/es/alimentacion/estadisticas. Marques RC (2010) Regulation of water and wastewater services. An international comparison. IWA Publishing, London. Martínez-Espiñeira R, García-Valiñas MA, González-Gómez F (2009) Does private management of water supply services really increase prices? An empirical analysis in Spain. Urban Stud 46(4):923–945. OECD (2004) Competition and regulation in the water sector. DAFFE/COMP(2004) 20. Organisation for Economic Co-operation and Development, Paris. Available online: http://www.oecd.org/regreform/sectors/ 33691325.pdf. Picazo-Tadeo AJ, Sáez-Fernández FJ, González-Gómez F (2009) The role of environmental factors in water utilities' technical efficiency. Empirical evidence from Spanish companies. Appl Econ 41(5): 615–628. Pigeon M, McDonald DA, Hoedeman O, Kishimoto S (eds.) (2012) Remunicipalisation. Putting water back into public hands. Transnational Institute, Amsterdam. Available online:http://www.tni.org/sites/www.tni.org/ files/download/remunicipalisation_book_final_for_web_0.pdf. Piriou P, Devesa R, Puget S, Thomas-Danguin T, Zraick F (2014) Evidences of regional differences in chlorine perception by consumers: sensitivity differences or habituation? J Water Supply Res T In Press, doi: 10. 2166/aqua.2014.097.

Private Water Utilities and User Perception of Tap Water Quality Romano G, Guerrini A, Vernizzi S (2013) Ownership, investment policies and funding choices of Italian water utilities: an empirical analysis. Water Resour Manag 27(9):3409–3419. Romano G, Salvati N, Guerrini A (2014) Factors affecting water utility companies' decision to promote the reduction of household water consumption. Water Resour Manag 28(15):5491–5505. Romano G, Masserini L, Guerrini A (2015) Does water utilities' ownership matter in water pricing policy? An Analysis of Endogenous and Environmental Determinants of Water Tariffs in Italy Water Pol In Press doi:10. 2166/wp.2015.016 Ruester S, Zschille M (2010) The impact of governance structure on firm performance: an application to the German water distribution sector. Util Policy 18(3):154–162. Scott J, Freese J (2014) Regression models for categorical dependent variables using stata (3rd edition). Stata Press, College Station. Turgeon S, Rodriguez MJ, Thériault M, Levallois P (2004) Perception of drinking water in the Quebec city region (Canada): the influence of water quality and consumer location in the distribution system. J Environ Manag 70(4):363–373. Wallsten S, Kosec K (2008) The effects of ownership and benchmark competition: an empirical analysis of US water systems. Int J Ind Organ 26(1):186–205. WHO (2011) Hardness in drinking-water. Background document for development of WHO guidelines for drinking-water quality. World Health Organization, Geneva Wooldridge JM (2002) Econometric analysis of cross section and panel data. The MIT Press, Cambridge Yoo S (2005) Analysing household bottled water and water purified expenditures: simultaneous equation bivariate tobit model. Appl Econ Lett 12(5):297–230