Abstract
We study the willingness to pay (WTP) for a large set of improvements in water service related to quality, continuity, and securing access for people with no house piped water during the COVID-19 pandemic. Using primary survey data from urban Peru, and the contingent valuation method, we estimate a mean WTP of around PEN 4.3 (USD 1.05), 3.7 and 1.8, respectively, for the aforementioned sets of improvements, with the combined WTP representing a 23% increase in the households’ water service monthly bill. The WTP for all sets of improvements is influenced by the expenditure in bottled water (a substitute for tap water, generally perceived as unsafe) and a proxy for household assets. The influence of the individual characteristics typically scrutinized by the literature (e.g., sex, age, and education) varies with the type of improvement examined. We find a significant heterogeneity in WTP across providers and calculate the users’ contribution to a water fund that could crowd-in the public investment in water services’ upgrades. We further discuss the implementation of such water fund.
Similar content being viewed by others
Notes
This figure was 39% in 2000, 34% in 2010, and 31% in 2015 (data from https://ourworldindata.org/clean-water, based on information from WHO/UNICEF Joint Monitoring Programme for Water Supply, Sanitation and Hygiene (JMP). Accesed: September 20, 2023).
Still, 10.6% of the population (or 3 million people) lack access to water. The average water access masks a substantial regional heterogeneity, with some regions attaining much lower access rates (e.g., Loreto in the Jungle: 63.6%, and Puno in the Highlands: 76.5%), a disparity that is exacerbated by the significant regional variation in the quality of the service they receive (MVCS 2021).
We are aware that, as in other emerging economies, rural areas in Peru have larger gaps in access to water (24% of the households versus 5% in urban areas). However, we do not have large scale survey information for rural areas. While the estimation of the WTP for improvements in water services would be important to assess how much water is valued, it is likely that the WTP figures would be smaller and the investments (for improvements in the current water services and new infrastructure) would be larger than in urban areas. If this were the case, the discussion from a policy perspective would be around the scope of a subsidy from the public sector. The study of this topic certainly deserves further research.
Appendix Table C1 reports the number of characteristics in water (and sewerage) service examined by a sample of studies for urban areas in developed and developing countries.
As of 2020, the density of complaints per 1000 connections went from 1 to 299, with an average of 103 and a standard deviation of 73. The average density of complaints has only been worse in the previous 5 years.
In terms of the number of connections managed, the figures are as follows: SEDAPAL, more than 1 million; each very large EPS, between 100,000 and 1 million; each large EPS, between 40,000 and 100,000; each medium-sized EPS, between 15,000 and 40,000; and each small EPS, fewer than 15,000. As of 2020, the total number of connections managed by those providers are: SEDAPAL, 1.54 million; very large EPS, 0.85 million; large EPS, 0.84 million; medium-sized EPS, 0.35 million; and small EPS, 0.10 million.
In rural areas, with fewer than 2000 inhabitants, WSS are provided by more than 25,000 Sanitation Services Administrative Boards (JASS, for its acronym in Spanish) and other providers.
The PMO includes a plan of investments describing which projects should be carried out, and how they will be financed. The time horizon of the PMO is up to 30 years (SUNASS 2020).
While we could have considered mail or internet-based interviews, it is not clear that any of those modes is superior to telephone interviews for Peru because, the response to unsolicited postal o electronic mails is particularly low, and the Internet access is not as extended as the telephone’s. Maguire (2009) mentions three sources of biases that could affect CV surveys: social desirability (the intent to appear in a more favorable fashion in the presence of the interviewer), avidity (those who are more interested in the survey topic are more likely to respond) and non-response (the composition of the sample that chose to complete the survey). The first one could affect perhaps even more strongly to in-person rather than telephone surveys. Moreover, given the topic (water), we could expect avidity to trigger in both in-person and telephone interviews (in Sect. 5, page 16, we provide evidence suggesting the absence of a significant overstatement of WTP due to the greater need for safe water in COVID-19 times). Lastly, the average response rate we got was around 60%, which is slightly smaller than the response rates the survey company in charge of the data collection registered in in-person interviews before 2019 (around 70%). From the calls that were accepted, 98.4% were made to cell phones, and the remaining 1.6% were made to land line numbers.
The geopolitical division of Peru includes 25 regions (akin to a US State), 196 provinces, and 1874 districts, as of December 2019. Another important distinction made when analyzing the data is among the three natural regions in Peru: The Coastal area (Costa), bordering the Pacific Ocean; the Highlands (Sierra), which is a section of the South American Andes; and the Jungle (Selva), the Peruvian section of the Amazon.
See: https://www.osiptel.gob.pe/media/jokj0o1g/np24052022-lineas-moviles.pdf. Visited on July 1, 2023.
The data collection was in charge of an experienced company conducting socioeconomic (in-person and telephone) surveys in Peru. Our research team trained all the pollsters. During the training, we paid special attention to explain the rationale behind the contingent valuation single-bounded and double-bounded questions, so that they could appropriately collect that information and respond to any questions from respondents. A set of pilot surveys was conducted to test the software, procedures, and clarity of the instructions, as well as to time the length of the questionnaire. The sampling errors are 2% (for SEDAPAL users), 3.7% for Very Large EPS, 6.3% for Large EPS, 11.2% for Medium-sized EPS, and 18.8% for Small EPS. Once we had defined the sample size, users from each type of provider were randomly selected until the desired number was reached.
One could argue that this variable could not properly capture the tenancy of assets in the Sierra and the Selva. However, the percent of households reporting their houses wall to be made with brick and mortar is fairly high in all three natural regions: 89% on the Coastal region, 74% in the Sierra, and 70% in the Selva.
Those participants who perceived the tap water as unsafe are significantly more likely to buy bottled water (50.9% vs. 41.2%) and to spend more in that purchase (PEN 13.00 vs. 10.86) than those who did not. The p-values from the respective (one-tail) tests of the mean differences are 0.0004 and 0.0244. In addition, 93% percent of respondents in our sample make some treatment to tap water (mostly boiling) before drinking it.
The figures from the 2020 SUNASS report mentioned earlier are close to the ones reported by our survey: 21.4 (SEDAPAL), 16.4 (very large EPS), 15.5 (large EPS), 19.7 (medium-sized EPS), and 14.3 (small EPS).
In the case of Peru, we know very little about the WTP for water in pandemic times. One exception is Gómez-Lobo et al. (2022), who compare households connected and non-connected to a water network in Lima, and find that the former households were less likely to report COVID-19 infections than the latter. We found no study on WTP for water during the COVID-19 pandemic, however.
In the entire sample, 55% of respondents either completed post-secondary technical education or higher. Those figures are 60.3% for users from SEDAPAL, 75.8% for users from very large EPS, 66.2% for users from large EPS, 56.9% for users from medium-sized EPS, 48.6% for users from small EPS, and 33.9% for users from UGM.
As part of our research project, we conducted 42 focus groups with about 336 WSS users from all types of providers, to identify the problems with the service and the amounts that users would be willing to pay for the related improvements. The set of bids used in our CV study roughly correspond to those amounts, in particular the maximum.
Two alternative methods of non-market goods valuation include choice experiments and travel cost. Carson et al. (1996) conducts a meta-analysis comparing CV values with those from travel cost values. In general, the authors find lower CV values, which we can take as a lower bound, at least compared to the travel cost method. Furthermore, designing a field experiment that compares real payments with stated hypothetical WTP elicited with a cheap talk, Blumenschein et al. (2008) finds that using a follow-up question removes the hypothetical bias in the CV method.
An interesting question to examine would be whether the WTP for water quality and continuity remains when the third group is not water access for the poor but any other improvement. We defer this to future research.
As part of our research project, we conducted 42 focus groups with about 336 WSS users from all types of providers. The bids used in our survey roughly respond to the values stated by those users; in particular, the maximum amount (PEN 11).
The 3-group sequence generated for each respondent took this form: “xyz”, where the values 123, 231, and 312, which reflect the order of each group, were equally likely. For instance, a respondent with the sequence 3,1,_,_,_,2 received CV questions for group 2, group 6, and group 1, in that order. The blanks mean this subject did not receive questions for groups 3, 4, and 5.
In each of the three groups of WTP questions, the respondents were told that they will be asked three times (i.e., for three groups of improvements), so that, when answering, they should consider that the implementation of those improvements would reduce their disposable income by the aggregate amount they stated. See Appendix B.
The questionnaires used for the other groups of improvements in water service examined are available from the authors upon request.
The results are similar when we estimate Probit models. Available upon request from the authors.
93% percent of our sample makes some treatment to tap water before drinking and 83% of the sample perceives the tap water as unsafe. This figure is similar across providers. Also, 50% of the sample buys bottled water regularly.
If we included income in the specification, we would see that this variable is positively correlated with WTP (though we lose a significant number of observations due to non-responses). Results are available upon request.
Although we use the same base specification for all sets of improvements in water service, our results are robust to alternative (more complete) specifications, as we show in Sect. 5.1.
We should be cautious with this interpretation, since the WTP figures across groups of improvements are not strictly comparable: the former measures improvements in quality (e.g., going from water with particles to crystal clear), while the latter asks for improvements in quantity (continuity), no interruptions in service and adequate pressure.
We assert that the WTP figures somewhat capture altruism, because they are not affected by who the provider is, unlike with happens with water quality and continuity.
We further added a variable capturing the households’ perception of the importance of water to fight COVID-19, collected in our survey, in the logit bid regression, under the premise that if households think that water was particularly important during that period (as is the case; see last row in Table 1), this could be reflected in their WTP. In none of the cases did this variable resulted statistically significant (results available from the authors).
The aforementioned figures assume a vertical growth in water access (i.e., no new connections).
Since the quality in the provision of the service considers different attributes (e.g. water continuity, pressure, density of breaks in the water network, density of blockages in the sewage network, among others), this figure represents only a portion of the quality gap. Thus, the indicators considered were: the percentage of the population with access to a safe water service and that with access to a safe sanitation service (from the World Bank’s World Development Indicators). ‘Safe water’ is defined as accessible water, available when needed (i.e., 24 h a day), and free of any contaminant. ‘Safe sanitation’ is defined as access to sanitation facilities not shared with other households, by which excreta are safely disposed of, on-site or transported, and subsequently treated.
Estimating a logit regression using the specification with the regular controls and adding the variable “certainty with the response to the CV question”, we find that the provider fixed effects were statistically significant for groups 1 and 2, whereas satisfaction with the WSS and certainty with the response to the CV question were significant for group 6 (unreported results). For the heterogeneity analysis, we compare users highly satisfied and highly certain with their responses to their respective counterparts.
References
Ahsan N, Hadiujjaman S, Islam S, Nasrin N, Akter M, Ara Parvin G, Hossain S (2021) Willingness to pay for improved safe drinking water in a coastal urban area in Bangladesh. Water Policy 23:633–653
Amoah A, Moffatt PG (2021) Willingness to pay for reliable piped water services: evidence from urban Ghana. Environ Econ Policy Stud 23:805–829
Blumenschein K, Blomquist GC, Johannesson M, Freeman P (2008) Eliciting willingness to pay without bias: evidence from a field experiment. Econ J 118:114–137
Bonifaz JL, Urrunaga R, Aguirre J, Quequezana P (2020) Brecha de infraestructura en el Perú: estimación de la brecha de infraestructura de largo plazo 2019–2038. [Coordinators: Pastor, C., and Brichetti, J. P.], Monograph prepared for the Inter-American Development Bank (IDB; 838)
Carson RT, Flores NE, Martin KM, Wright JL (1996) Contingent valuation and revealed preference methodologies: comparing the estimates for quasi-public goods. Land Econ 72:80–99
Chatterjee Ch, Triplett R, Johnson CK, Ahmed P (2017) Willingness to pay for safe drinking water: a contingent valuation study in Jacksonville, FL. J Environ Manag 203:413–421
El Peruano (2017) Official Newspaper. Lima, Peru
Felgendreher S, Lehmann P (2016) Public choice and urban water tariffs—analytical framework and evidence from Peru. J Environ Dev 25(1):73–99
Fujita Y, Fujii A, Furukawa S, Ogawa T (2005) Estimation of willingness to pay (WTP) for water and sanitation services through contingent valuation method (CVM): a case study in iquitos City, The Republic of Peru. JBICI Rev 11:59–87
Gómez-Lobo A, Gutiérrez M, Huamaní S, Marino D, Serebersky T, Solís B (2022) Access to water and COVID-19: a regression discontinuity analysis for the peri-urban areas of Metropolitan Lima, Peru. IDB Working Paper 01351
Khan NI, Brouwer R, Yang H (2014) Household’s willingness to pay for arsenic safe drinking water in Bangladesh. J Environ Manag 143:51–161
Krinsky I, Robb AL (1986) On approximating the statistical properties of elasticities. Rev Econ Stat 68(4):715–719
Lopez-Feldman A (2010) DOUBLEB: Stata module to estimate contingent valuation using Double-Bounded Dichotomous Choice Model
Maguire (2009) Does mode matter? A comparison of telephone, mail, and in-person treatments in contingent valuation surveys. J Environ Manag 90(11):3528–3533
Makwinja R, Kosamu IBM, Kaonga ChCh (2019) Determinants and values of willingness to pay for water quality improvement: insights from Chia Lagoon, Malawi. Sustainability 11:4690
MVCS (2021) Plan Nacional de Saneamiento 2022–2026. Lima: Ministerio de Vivienda, Construcción y Saneamiento (MVCS). https://www.gob.pe/institucion/vivienda/normas-legales/2586312-399-2021-vivienda. Visited: September 10, 2022
MEF (2021) Nota Técnica para el uso de los Precios Sociales en la Evaluación Social de Proyectos De Inversión. Lima: Ministerio de Economía y Finanzas (MEF)
Instituto Nacional de Estadística e Informática—INEI (2021a). Perú: Estado de la Población en el año del Bicentenario, 2021. Lima: INEI
Instituto Nacional de Estadística e Informática—INEI (2021b) Estadísticas de las Tecnologías de Información y Comunicación en los Hogares. Lima: INEI
OECD (2021) Water Governance in Peru. OECD Studies on Water. OECD Publishing, Paris
Olmstead SM, Stavins RN (2009) Comparing price and nonprice approaches to urban water conservation. Water Resour Res 45:W04301
OSIPTEL (2021) Encuesta Residencial de Servicios de Telecomunicaciones 2021. Available at: https://repositorio.osiptel.gob.pe/handle/20.500.12630/327 (Visited on 28 June 2023)
Rodríguez-Tapia L, Revollo-Fernández DA, Morales-Novelo JA (2017) Household’s perception of water quality and willingness to pay for clean water in Mexico City. Economies 5(2):12
Rogers P, de Silva R, Bhatia R (2002) Water is an economic good: how to use prices to promote equity, efficiency, and sustainability. Water Policy 4:1–17
Soto Montes de Oca G, Bateman IJ (2006) Scope sensitivity in households’ willingness to pay for maintained and improved water supplies in a developing world urban area: Investigating the influence of baseline supply quality and income distribution upon stated preferences in Mexico City. Water Resour Res 42:W07421
SUNASS (2020) Agua Potable y Saneamiento. Más esenciales que nunca. Memoria 2020. https://www.sunass.gob.pe/sunass-te-informa/publicaciones. Visited: September 10, 2022
Tanellari E, Bosch D, Boyle K, Mykerezi E (2015) On consumers’ attitudes and willingness to pay for improved water quality and infrastructure. Water Resour Res 51:47–57
Tudela-Mamani JW, Leos-Rodríguez JA, Zavala-Pineda MJ (2018) Estimation of economic benefits for improvements in basic sanitation services using the contingent valuation method. Agrociencia 52:467–481
Tussupova K, Berndtsson R, Bramryd T, Beisenova R (2015) Investigating willingness to pay to improve water supply services: application of contingent valuation method. Water 7(6):3024–3029
United Nations—UN (2010) Resolution Adopted by the General Assembly on 28 July 2010. A/RES/64/292
Van Houtven GL, Pattanayak SK, Usmani F, Yang J-C (2017) What are households willing to pay for improved water access? results from a meta-analysis. Ecol Econ 136:126–135
Vásquez WF, Espaillat R (2016) Willingness to pay for reliable supplies of safe drinking water in Guatemala: a referendum contingent valuation study. Urban Water J 13(3):284–292
Vásquez WF, Mozumder P, Hernández-Arce J, Berrens RP (2009) Willingness to pay for safe drinking water: evidence from Parral, Mexico. J Environ Manag 90(11):3391–3400
Venkatachalam L (2004) The contingent valuation method: a review. Environ Impact Assess Rev 24:89–124
World Bank (2018) The World Bank Modernization of Water Supply and Sanitation Services (P157043). https://documents1.worldbank.org/curated/en/706771525142348131/text/Project-Information-Document-Integrated-Safeguards-Data-Sheet-Modernization-of-Water-Supply-and-Sanitation-Services-P157043.txt. Visited: September 12, 2022
World Health Organization (WHO) and the United Nations Children’s Fund (UNICEF) (2021) Progress on household drinking water, sanitation and hygiene 2000–2020: five years into the SDGs. WHO and UNICEF, Geneva
Acknowledgements
Francisco Taquiri provided valuable research assistance. We thank comments from seminar participants at the Peruvian Economics Association Annual Meeting, Universidad del Pacífico (Viernes CIUP), Universidad de Buenos Aires (Jornadas Argentinas de Econometría), LACEA, and BCDE. We are grateful to two anonymous referees for valuable feedback on an earlier version of the manuscript. The primary survey data used in this paper were collected as part of a project in which José Luis Bonifaz was part of the team. We thank the financial support from SECOSAN Program of the Swiss Cooperation in Peru, made up of the Swiss State Secretariat for Economic Affairs (SECO), the Peruvian Ministry of Housing, Construction and Sanitation, the National Superintendence of Sanitary Services (SUNASS), and the Technical Agency of the Administration of Sanitation Services (OTASS). SECOSAN did not participate in the design, implementation, data collection, or analysis of this paper. All errors are our responsibility.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
About this article
Cite this article
Galarza Arellano, F.B., Carbajal, M. & Aguirre, J. Willingness to pay for improved water services: evidence from Peru. Environ Econ Policy Stud 26, 503–539 (2024). https://doi.org/10.1007/s10018-023-00381-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10018-023-00381-1