Water and. Sanitation Program. Water and. Sewerage Systems

Transcription

1

2

3 Water and Sanitation Program Water and Sewerage Systems

4 Report written by: Vivien Foster The Word Bank. Primariy thanks to the peer reviewers that have contributed with vita comments and suggestions to this report: Caroine van den Berg, Barbara Evans, Agnieska Grudzinska and Carmen Arévao from the Water and Sanitation Program; and Kaus Neder from CAESB, Brazi. Aso thanks to the E Ato Piot Project team: Luiz Lobo, Regina Arakaki, Mery Quitón, speciay to Neba Cannei and Adea Martínez; and the WSP-AND team: Joseph Narkevic, Louise Herrmann and Ann-Katrin Petersen. A specia mention to Isabee Vincent and Luis Curridor from the concessionaire Aguas de Iimani and Pedro Arrioa and Caudio Parra from Caja Los Andes. Besides this, the WSP-AND pubication coud not have been done without the coaboration from the Vice Ministry of Basic Services in Boivia, Water Superintendence in Boivia and Prefecture of La Paz. Lasty, and perhaps most importanty, thanks to the Swedish Internationa Deveopment Cooperation Agency, SIDA, main financia contributor to the E Ato Piot Project. For information contact: Water and Sanitation Program (WSP) Word Bank Office, Lima Ávarez Caderón 185, 9th foor San Isidro, Lima-Perú Phone: (511) Fax: (511) E-mai: [email protected] Website: Design: Fabioa Pérez Abea Edition: Santiana S.A. Printed in Peru ISBN-Nr.:

5 Water and Sanitation Program EL ALTO - BOLIVIA / PILOT PROJECT Water and Sewerage Systems Costs of Impementation of the Mode Viceministerio de Servicios Básicos Agencia Sueca para e Desarroo Internaciona ECONOMIC AND FINANCIAL EVALUATION

6

7 EL ALTO - BOLIVIA / PILOT PROJECT CONTENTS Executive Summary 6 1. Introduction Description of the E Ato Piot Project 12 Project Setting Timing of the Piot Project Services Provided Condominia Approach 3. Evauation Methodoogy 18 Condominia versus Conventiona Different Anaytica Perspectives 4. Cost Differentias 24 Financia Costs Economic Costs 5. Consumption Differentias 38 Estimating the Consumption Differentia Vauing the Consumption Differentia 6. Evauation 46 Socia Perspective Utiitiy s Perspective Consumer s Perspective 7. Concusions 50 Cost Annex 58

8 WATER AND SANITATION PROGRAM Executive Summary This paper undertakes the economic and financia evauation of the E Ato Piot Project (EAPP), a piot project aimed at transferring the condominia water and sewerage system from Brazi to Boivia and testing its appicabiity in the context of private sector participation in service provision. Since its inception in 1998, the EAPP has provided condominia water connections to 1,977 househods in eight neighborhoods of E Ato, and condominia sewerage connections to 4,050 househods in nine neighborhoods of E Ato. According to recent statistics, about 60% of these househods ive beow the poverty ine with an average daiy income per capita of US$0.80. To what extent are the resuts of the E Ato piot reevant beyond Boivia? The pecuiar cutura, geographica and socia circumstances of E Ato make it if anything an acid test for the condominia approach. In particuar, a number of factors which have served to imit the benefits of the condominia system in Boivia woud not necessariy be present in other contexts to the same degree. Exampes incude the exceptionay ow eves of consumption of househods, and the difficuties experienced in inducing them to switch to modern hygiene practices for a variety of cutura reasons. Consequenty, the resuts of the evauation shoud be regarded as specific to the E Ato context, even though in quaitative terms they are indicative of what can generay be achieved through the condominia approach. The piot project experimented with a number of different components incuding: - innovative engineering design of networks; - community participation in network construction and maintenance; - hygiene education to support the instaation of househod faciities; - micro-credit ines to finance the construction of bathrooms. Many of these components are mutuay reinforcing and formed part of an integrated concept in the E Ato project. Nonetheess, both in principe as in practice, it is possibe to appy these components independenty of each other. Hence, for anaytica purposes it proves convenient to provide an independent evauation of each. Engineering design The purpose of the innovative engineering design is to reduce the ength, diameter and depth of the network required by routing the distribution pipes across pavements and/or backyards. Anaysis of the EAPP experience suggests that savings in the ength and diameter of pipes are of the order of 10%-20%, whie savings in the voume of soi excavation as a resut of shaower trenches are of the order of 45%-75%. These 6

9 EXECUTIVE SUMMARY physica savings transate into overa financia savings of the order of 24% for the sewerage service and 40% for the water service when the condominia engineering design is impemented using conventiona contractors. This is consistent with recent experience from Brasiia which suggests savings of 20% for condominia sewerage systems without community participation (Neder, 2001). Community participation Community participation brings a number of advantages, among them is a further reduction in connection costs as a resut of training oca residents to construct and maintain their own condominia branches. Community participation reduced the network costs by a further 26% for the sewerage service and 10% for the water service. Thus, the overa savings achievabe by impementing the condominia engineering design with community abor come to around 50% for each of the two services. However, community participation aso introduces costs of socia intermediation for the water company of around US$8 per connection, and requires each participating househod to give-up about a week of its time vaued at around US$20. When these costs are taken into account, the overa cost advantage of the condominia design with community participation fas sighty from 50% to 40% for both services. Considering the baance between margina costs and benefits, community participation has a benefit cost ratio of 2.2 for the sewerage service suggesting that it is very worthwhie. However, in the case of the water service the benefit cost ratio is ony 1.0, since in this case the cost savings come primariy from the engineering design. There is aso some evidence suggesting that community participation increases the proportion of househods that connect to the sewerage network once it is buit from 66% to 75%. The effect of this is to increase the overa cost saving of the approach from around 40% to 45%. Hygiene education The purpose of the hygiene education component was to provide mora and technica support for househods to adopt modern hygiene practices, in particuar by heping them to construct their own bathrooms and associated faciities. Without such investments within the home, a sewerage connection brings itte or no benefit to househods, and has been shown to have virtuay no impact on water consumption. In neighborhoods where hygiene education activities had been performed the probabiity that any given househod had constructed a bathroom increased from 38% to 73%, and the range of water-using instaations in the househod (showers, kitchen sinks, etc.) increased markedy. Househods with bathrooms were found to consume, on average, two cubic meters more per month than househods without bathrooms, which is to say an increase of 30% on average consumption in househods without bathrooms of 5.4 cubic meters per month. Househods receiving hygiene education were aso found to significanty reduce insanitary practices (Canei, 2001). After the project, the percentage of househods throwing out used water into the streets fe from 77% to 58%. Househods recycing water within the home fe from 36% to 25%. The cost of hygiene education to the utiity works out at around US$13 per househod, whie the cost of bathroom construction to the househod was estimated at US$443. Given the reativey sma absoute vaue of the consumption increase, and the ow eve of tariffs for ow voume consumers, hygiene education is not at present a commerciay attractive proposition for a private utiity in E Ato. The estimated consumer surpus increase from the additiona consumption ranges from US$253-US$470 depending on the discount rate used. This suggests that bathroom construction is ony marginay attractive to househods. However, it is important to stress that these modest benefits for hygiene education are argey an artifact of the exceptionay ow eves of water consumption in E Ato. With the much arger consumption increases that coud easiy be anticipated in other ocations, the proposition woud start to ook much more attractive both from the utiity s and the customer s perspectives. EL ALTO - BOLIVIA / PILOT PROJECT 7

10 WATER AND SANITATION PROGRAM Micro-credit ine The purpose of the micro-credit ine was to hep househods finance the US$400 worth of materias required to construct a fuy-equipped bathroom. Overa, 25% of househods appied for credit, and 19% had their appications approved. The avaiabe evidence suggests that those househods appying for credit tended to have above average incomes. Whie the provision of credit undoubtedy eased the payment of the costs for beneficiary househods, statisticay there is no evidence to suggest that bathroom construction rates were significanty higher in neighborhoods were micro-credit had been offered. The microcredit ine was dropped at an eary stage of the project. Overa, these resuts suggest the foowing concusion. The engineering design component aone produced significant savings of 24% for sewerage and 40% for water. The community participation component further increased the savings avaiabe for the sewerage service to 40%, but did not have any net effect on the cost reductions avaiabe for the water service. The hygiene education component had a very substantia effect on househod behavior, but the size of the benefits was more modest than might be expected owing to very abstemious water consumption practices pecuiar to E Ato. Evidence on the micro-credit faciity suggests that it probaby was not a that effective in reaching the poorest househods. The resuts reported above are given from the perspective of an enightened poicy maker who takes a financia and economic costs fuy into account. However, the considerabe divergence between tariff structure and underying cost structures in the AISA concession contract appears to distort the way in which the utiity and its consumers perceive the advantages of the condominia system. From the utiity s perspective, it is difficut to make categorica statements about the profitabiity of making condominia as opposed to conventiona connections. Athough current connection charges mandated by the concession contract are substantiay ower than the estimated average cost of making a connection, it is not known to what extent part of the costs of network expansion have in fact been buit into the use of service tariff. However, one point that is cear is that there does not appear to be any incentive for the utiity to engage in hygiene education since the additiona costs are not recouped through the connection charge, and the resuting increases in water consumption though significant are not arge enough to take consumers out of the oss-making ow voume tariff band. If charges were fuy cost-refective the utiity woud be indifferent between conventiona and condominia systems, but woud sti ony have an incentive to undertake hygiene education as ong as the consumption increase was arge enough to justify the cost of the initia investment. From the consumer s perspective, househods receiving water and sewerage services by the condominia method saved US$19 and US$80 respectivey in terms of connection charges, an aggregate vaue equivaent to 80% of monthy househod income. However, when the opportunity cost of the househod s time is fuy accounted for, the saving disappears for the water service and for the sewerage service is reduced to US$58 (or about 50% of monthy househod income). If connection charges were more cosey aigned with underying cost, househods woud spend more on obtaining a condominia connection than they do at present, but the savings reative to the conventiona system woud aso be arger. They woud increase to US$109 for water and US$126 for sewerage (together equivaent to neary 200% of monthy househod income). As before, when the opportunity costs of time are taken into account, the savings are somewhat reduced to US$90 for water and US$104 for sewerage (in tota equivaent to 160% of househod income). 8

11

12 WATER AND SANITATION PROGRAM 1. Introduction The purpose of this report is to conduct an economic and financia evauation of the E Ato Piot Project (EAPP) for condominia water and sewerage systems in Boivia. The report does not however directy consider the technica performance of the systems, or the socia impacts of the projects which are being treated in parae evauations. Condominia water and sewerage systems were pioneered in Brazi during the 1980s as a way of bringing piped sanitation services within the economic reach of poor househods (Watson, 1995). The condominia approach has two defining features. The first is the use of innovative engineering techniques. By routing water and sewerage networks across pavements and yards instead of down the center of streets, the condominia approach eads to substantia economies in the ength, depth and diameter of the pipes. The second is the integration of socia and engineering work. By invoving communities in the construction and maintenance of the condominia networks, further cost reductions are achieved. Moreover, the interaction with the community during the execution of the works provides opportunities to impart hygiene education and to infuence water consumption habits in a variety of ways. In Juy 1997, the Government of Boivia granted a 30 year concession contract for the provision of water and sewerage services to the adjacent cities of La Paz and E Ato. The concession contract was awarded to Aguas de Iimani (AISA), a consortium ed by Lyonnaise des Eaux. A major objective of the concession contract was to improve access to water and sewerage services in E Ato. The expansion targets for the first four years of the concession in E Ato incuded reaching 100% water coverage and making 38,000 new sewerage connections. However, the very high eves of poverty in E Ato raised concerns about the affordabiity of the new water and sewerage services. This prompted a search for ways to reduce the cost of serving ow income househods, and ed the Boivian authorities to consider the potentia use of the condominia approach. Shorty after the award of the concession contract in Juy 1997 a tripartite agreement was reached between the Government of Boivia (GoB), the private utiity (AISA) and the Water and Sanitation Program (WSP). 10

13 1. INTRODUCTION - The GoB agreed to reax its technica standard which, as currenty written, woud egay precude the use of the condominia approach. - With the endorsement of the reguatory agency, AISA agreed to use the condominia approach to meet a proportion of its expansion targets in E Ato. - Whie the WSP thanks to support from the Swedish Internationa Deveopment Cooperation Agency (SIDA) agreed to finance the research and training activities required to transfer and adapt the condominia system from Brazi to Boivia. Athough condominia systems have been successfuy depoyed in a number of cities around Brazi, the EAPP is beieved to represent the first attempt to export the approach to another country with very different socioeconomic and sectora conditions. Furthermore, the Braziian experience of condominia systems has aways been in the context of pubic sector service provision. The EAPP appies the mode, for the first time, in the context of private sector participation. It therefore provides an opportunity to estabish whether the socia work component of the methodoogy is compatibe with the modus operandi of a private utiity. For both of these reasons, the experience is of interest not ony in Boivia, but to other countries who may wish to experiment with the condominia system. This report is organized as foows. - Section 2 provides a more compete description of the EAPP. - Section 3 describes the methodoogica framework for the evauation. - Section 4 estimates the cost differences between conventiona and condominia systems. - Section 5 examines how the condominia approach affects househod water consumption. - Section 6 draws out the main concusions of the evauation. EL ALTO - BOLIVIA / PILOT PROJECT 11

14 WATER AND SANITATION PROGRAM 2. Description of the E Ato Piot Project This section provides a more detaied description of the EAPP project, describing the setting in which it takes pace, the nature and phasing of the neighborhoods affected by the project and the scope of the associated interventions. It concudes by providing a more nuanced definition of what is meant by the condominia approach. Project setting The setting for the EAPP is the city of E Ato ocated at 4,100 meters above sea eve on the Boivian atipano 1, adjacent to La Paz and home to 600,000 peope. The city has been described as the Aymara capita of the word, in recognition of the fact that 80% of its popuation beong to the Aymara ethnic group, and that Aymara beiefs and customs are strongy observed. The city has grown-up from nothing during the ast 30 years (and more especiay during the ast 10 years) as a resut of arge scae urbanrura migration. However, many of the inhabitants retain strong inks with the countryside, spending a proportion of each year back in their viages where they continue to take part in agricutura activities. For a considerabe proportion of the popuation, E Ato thus represents a way of combining urban and rura iveihoods. Whie in the city, most peope earn their iving by street vending or informa micro-enterprises, as there is very itte organized arge scae industry. Many inhabitants of E Ato spend most of their waking hours in La Paz where most of the economic opportunities are, hence E Ato is something of a dormitory city. The recent creation and rapid growth of the city has created a coverage deficit, particuar for sanitation. Sewerage coverage in E Ato, estimated at 30% - 45%, ags significanty behind that in La Paz, estimated at 66%. However, the high eves of poverty make expansion of services particuary chaenging: average income in E Ato is ess than haf of that in La Paz, with 60% of the popuation iving under the poverty ine (Tabe 2-1). Tabe 2-1 Poverty statistics for La Paz and E Ato La Paz E Ato Income (US$/month) Per capita Per househod Poverty ines (US$/per capita/month) Extreme poverty Poverty Poverty rates (%) Extreme poverty Poverty

15 2. DESCRIPTION OF THE EL ALTO PILOT PROJECT A of these characteristics combine to make E Ato a unique ocation, and raise a number of critica issues for the expansion of water and sanitation services (Tabe 2-2). Tabe 2-2 Aspect Loca idiosyncrasies of reevance to the project Loca characteristics Impications for water and sanitation services First, the reigious beiefs of the Aymara peope precude certain forms of modern hygiene. For exampe, atrines and septic tanks woud be regarded as unacceptabe to many because they invove the buria of feces in the ground, which is considered both dangerous and sacriegious. However, piped sewerage does not create confict with traditiona beiefs, basicay because it returns feces to surface water courses. Second, the geographic ocation means that temperatures are unusuay cod for a tropica ocation. This means that water use, in particuar for persona washing, is exceptionay ow; on average just under six cubic meters per househod per month or about 40 iters per capita per day. Furthermore, the cimate, ow oxygen eves and high eves of utravioet radiation create an unusuay sterie environment, so that the inks between hygiene and heath are not as strong as they usuay are in tropica ocations. Third, the argey rura origins of the popuation and the continued inks with the countryside, make many peope reuctant to adapt to modern urban ifestyes. Peope are accustomed to obtaining their water directy from nature, and returning their feces directy to nature. They are not used to iving with piped water and sewerage, and the oder peope in particuar are not aways attracted to doing so. Cutura Geographic Socia 80% of the popuation of E Ato beongs to the Aymara ethnic group, with strong traditiona reigious beiefs: - The earth mother (pachamama) is venerated as an important deity. - Evi spirits reside in the depths of the earth and can escape through hoes in the ground. - Possession of another person s feces gives power of witchcraft over them. E Ato is ocated at an atitude of 4,100m above sea eve, which means that: - Overnight temperatures fa to freezing year round. - Low temperatures, ow oxygen eves, and high eves of utravioet radiation combine to reduce bacteria activity in the oca environment. E Ato is a city that has been created in the ast 10 years as a resut of rura-urban migration, hence: - Latrines and septic tanks are not a cuturay viabe option for sanitation. - Buria of feces in the earth is regarded as sacriegious. - Creation of hoes in the ground for the disposa of feces is regarded as dangerous. - There is a preference for anonymous disposa of feces in scattered ocations. - Hot water is needed for showering. - Heath outcomes are ess sensitive to hygiene practices than they woud be in a warmer cimate. - Genera water use tends to be ower than in warmer cimates. The pecuiar cutura, geographica and socia circumstances of E Ato make it if anything an acid test for the condominia approach. In particuar, a number of factors which have served to imit the benefits of the condominia system in Boivia woud not necessariy be present in other contexts to the same degree. Exampes incude the exceptionay ow eves of consumption by househods, and the difficuties experienced in - Most of the residents are recent arrivas from rura areas. - Many of the residents maintain strong inks with the rura areas, dividing their time between agricutura activities in the countryside and commercia activities in the city. - Peope have imited experience of modern sanitation practices. - Difficuty of coordinating projects requiring agreement between neighbors, given seasona migrations and many absentee andords.

16 WATER AND SANITATION PROGRAM inducing them to switch to modern hygiene practices for a variety of cutura reasons. It is particuary important to bear these oca idiosyncrasies in mind when drawing inferences about the reevance of the E Ato experience to other ocations in Boivia, Latin America or beyond, where many of these conditions are unikey to appy. Timing of the Piot Project The EAPP encompasses a number of different neighborhoods of E Ato, as we as some of the poorer neighborhoods of La Paz (Caja Ferroviaria, E Rosa Loreta, and Kupini). The impementation of the project can be divided into two partiay overapping phases (Tabe 2-3). Phase One, which was the piot proper, ran from November 1998 to February 2000 and encompassed six different neighborhoods: Huayna Potosí, Via Ingenio, Caja Ferroviaria, San Juan de Río Seco, Oro Negro, Jichusirca. During this phase, a team of WSP consutants had overa responsibiity for the project sites. The team, which combined Braziian condominia experts with Boivian counterparts, undertook the design of the condominia networks, and the training of oca socia workers and AISA engineers, as we as supervision of the overa process. The Phase One sites were covered in two waves. The two projects covered in the first wave, Huayna Potosí and Via Ingenio, provided the basic earning experience. This is refected in the reativey ong duration of these projects, which took on average about eight months from start to finish. The four neighborhoods covered in the second wave, on the other hand, were executed much more rapidy, taking on average just under four months to compete. Phase Two, which ran from October 1999 to November 2000, was the first stage of scae-up. A key change between Phase One and Phase Two was that AISA began to take overa responsibiity for impementing the condominia methodoogy, whie the WSP team payed a much more imited advisory roe. Five further neighborhoods have been added under this phase: Via Ingenio (a different part of the origina Phase One neighborhood), E Ingenio, Germán Busch, E Rosa Loreta, and Kupini. These projects have taken, on average, just over five months each to compete. Services provided The nature of the services provided differs by piot neighborhood (Tabe 2-4), depending on their origina service endowment. Some of the piot neighborhoods Huayna Potosí, Oro Negro and Tabe 2-3 Phasing of piots by neighborhood o n d j f m a m j j a s o n d j f m a m j j a s o n d Phase One Huayna Potosí Via Ingenio (D-II) Caja Ferroviaria San Juan de Río Seco Oro Negro Jichusirca 199 days 279 days 168 days 105 days 119 days 74 days Phase Two Via Ingenio (UV-4) E Ingenio Germán Busch E Rosa Loreta Kupini II 113 days 203 days 160 days 175 days 130 days 14

17 2. DESCRIPTION OF THE EL ALTO PILOT PROJECT Kupini aready had a conventiona water suppy. Hence, the piot was imited to adding a condominia sewerage service. Esewhere, condominia water and sewerage systems were simutaneousy provided. The ony exception is Jichusirca, where ony water was provided. reasons, it may not aways be feasibe to confine the networks to the sidewaks. To the extent that the network has to cross through private yards, community participation becomes indispensabe because of the probems of access. Condominia approach The EAPP incorporated a number of different innovative components. Many of these components are mutuay reinforcing and formed part of an integrated concept in the E Ato project. Nonetheess, both in principe as in practice, it is possibe to appy these components independenty of each other. Hence, for anaytica purposes it proves convenient to provide an independent evauation of each. Engineering design component: This refers to a modification of conventiona design so that networks are routed aong sidewaks or private yards, rather than down the midde of roads. This aows for savings to be made in the ength, depth and diameter of pipes. If this approach is impemented directy by the utiity, without the community participation component, then networks must generay be confined to sidewaks to faciitate access during construction and subsequent access for maintenance. However, for topographica Community participation component: This means that the responsibiity for constructing and maintaining the branches ( ramaes ) of the condominia network is entirey deegated to the community. The operator empoys a team of socia workers to engage with the community and to train them for these activities. Community invovement heps to improve the acceptabiity of the infrastructure, promoting network connections, and provides an entry point for imparting hygiene education. Moreover, since the abor is provided free of charge, the financia costs of doing the works is reduced. However, against this must be set the additiona cost of the socia workers required to faciitate this process. Within topographica constraints, the community are aowed to choose whether the condominia network shoud pass aong the sidewak or through their private yards. It is important to note that, community participation does not necessariy entai routing through private yards, however routing Tabe 2-4 Characteristics of piot by neighborhoods Water Service Sewerage Service Before After Before After Phase One Huayna Potosí Conventiona Nothing Condominia Via Ingenio (D-II) Nothing Condominia Nothing Condominia Caja Ferroviaria Nothing Condominia Nothing Condominia San Juan de Río Seco Mixed* Mixed** Nothing Condominia Oro Negro Conventiona Nothing Condominia Jichusirca Nothing Condominia Nothing Phase Two Via Ingenio (UV-4) Nothing Condominia Nothing Condominia E Ingenio Nothing Condominia Nothing Condominia Germán Busch Nothing Condominia Nothing Condominia E Rosa Loejeta Nothing Condominia Nothing Condominia Kupini II Conventiona Nothing Condominia EL ALTO - BOLIVIA / PILOT PROJECT *Some househods had no water and others had conventiona water connections. ** Those with conventiona connections retained them and the others obtained condominia connections. 15

18 WATER AND SANITATION PROGRAM Tabe 2-5 Possibe permutations through private yards necessariy entais community participation (Tabe 2-5). In the case of EAPP ony the neighborhoods of Huayna Potosí and Via Ingenio (and partiay Caja Ferroviaria) have used yard routings of the condominia branches. In a other neighborhoods, the branches were routed across the sidewaks. Hygiene education component: The process of community engagement needed for the construction of condominia networks can aso be used to enter into a wider diaogue about water use. The nature of this diaogue is ikey to vary with the context. In the case of E Ato, a key issue was the very ow awareness of hygiene issues and imited exposure to modern sanitation faciities. Hence, considerabe emphasis was paced on teaching househods about the connection between water and heath, and providing them with mora and technica support for the With Community Without Community Network Routing Participation Participation Pubic sidewaks Possibe Possibe Private yards Possibe Impossibe Tabe 2-6 Permutations of the condominia approach by neighborhood Engineering Participation Education Credit Phase One Huayna Potosí ü ü ü ü Via Ingenio (D-II) ü ü ü ü Caja Ferroviaria ü ü ü San Juan de Río Seco ü ü ü Oro Negro ü ü ü Jichusirca ü ü Phase Two Via Ingenio (UV-4) ü E Ingenio ü ü Germán Busch ü ü E Rosa Loreta ü Kupini II ü construction of their own bathrooms and other househod faciities such as kitchen and aundry sinks. The education aso attempted to discourage undesirabe practices, such as disposing of waste water into the streets, or recycing it within the househod. Micro-credit faciity: This refers to the incorporation of a micro-credit faciity for the construction of bathrooms. The micro-credit was seen as a vauabe compement to the hygiene education, since there is itte point in encouraging househods to buid bathrooms if they cannot afford to do so. These four components are conceptuay distinct and can be packaged in a variety of different ways. What is commony referred to in the iterature as the condominia mode combines the engineering component with the community participation component. However, it is aso possibe to appy the engineering approach in isoation of the other interventions, whie hygiene education and micro-credit faciities can readiy be incorporated into a program invoving community participation. Under the EAPP, different combinations of these four components were tried in different neighborhoods (Tabe 2-6). During Phase One, a comprehensive approach was taken. A project sites incuded a community participation component 2, and educationa activities in support of bathroom construction were conducted everywhere except Jichusirca, where they were not reevant owing to the absence of sewerage service. In addition, a micro-credit ine was provided for househods in the first two piot sites at Huayna Potosí and Via Ingenio. Since the inception of Phase Two, the approach has changed consideraby. Most of the Phase Two sites have imited themseves to the engineering component 3. Ony in E Ingenio and Germán Busch has community participation been appied. The educationa component has aso argey been cut back. The reason for this was that the government s own socia investment fund (FNDR) is introducing a program to provide free bathrooms to househods in E Ato, making the educationa component as conceived in Phase One somewhat redundant. 16

19

20 WATER AND SANITATION PROGRAM 3. Evauation Methodoogy This section outines the evauation methodoogy that wi be used in this study. It begins by estabishing the conventiona system as the benchmark against which the reative costs and benefits of the condominia system wi be measured. The section goes on to define the various perspectives from which the evauation can be performed, and expains the reevance of each to the current exercise. Condominia versus conventiona In order to evauate an intervention such as the EAPP it is necessary to have a cear counterfactua, that is, to estabish what aternative system the condominia system is being compared against. In the context of E Ato, the reevant counterfactua is ceary the conventiona water and sewerage system, since AISA was contractuay obiged to connect the househods in the project area and woud have done so using conventiona technoogies had the EAPP not been undertaken. The evauation wi therefore concentrate on a comparison of the costs and benefits of condominia versus conventiona technoogies for water and sewerage provision. The impication of this is that it is ony necessary to focus on those costs and benefits that differ between the two systems. To this end, Tabe 3-1 highights those areas where differences arise between the two systems, and aso indicates how these differences vary according to the four components of the condominia approach identified in the preceding section. In many respects, the two systems are quite simiar. For exampe, the costs of water and sewage treatment, as we as the costs of biing and administration, are essentiay the same across the two systems and hence need not be considered in the evauation. Simiary, on the benefits side of the equation, the environmenta impact of the two systems is very simiar since this depends on the eve of sewage treatment appied which is not contingent on the nature of the upstream distribution network. However, key differences do arise in three areas: The first key difference is the cost of water and sewerage distribution networks. Reative to the conventiona approach, costs are reduced by appying the condominia engineering design to the distribution network. Further cost reductions are achieved by means of community participation, because the beneficiaries suppy their own abor for the construction of the condominia branches. Moreover, when it comes to operating costs of the distribution network, the condominia system reduces costs 18

21 3. EVALUATION METHODOLOGY Tabe 3-1 Comparison between conventiona projects and EAPP EAPP Engineering Participation Education Credit Conventiona Costs Capita expenditure (a) Water / sewage treatment No Difference (b) Water / sewerage distribution network Medium Low Low Low High Operating and maintenance costs (a) Biing and administration No Difference (b) Water / sewage treatment No Difference (c) Water / sewerage distribution network High Low Low Low High Socia intermediation Low Medium High High Low Benefits Consumption impact Low Low High High Low Environmenta impact No Difference to the extent that there is community participation in the maintenance of the condominia branches. The second area where the two systems differ is the cost of socia intermediation. Significant costs of socia intermediation ony begin to be incurred when community participation is introduced. These increase further if a component of hygiene education (and possiby micro-credit) is introduced. The important point to note is that the variations with the owest engineering costs aso tend to be those with the highest costs of socia intermediation. Hence, a trade-off is introduced and the key question becomes whether the additiona socia intermediation required to organize community participation is justified by the associated reduction in the costs of constructing and operating the network. The third key difference between the two systems is the impact on househod water consumption. Where hygiene education (and possiby micro-credit faciities) are introduced into the condominia package, it becomes possibe to have an impact on the construction of bathrooms and other househod water instaations, which in turn affects househod water consumption. This has both socia payoffs in the form of improved hygiene and financia payoffs in the form of higher utiity revenues. However, as before, the permutations with the highest consumption impact are aso those with the highest costs of socia intermediation. Once again, an important trade-off is introduced, namey whether the higher costs of socia intervention are justified in terms of the higher resuting eves of water consumption. The evauation exercise wi therefore focus on two aspects. - The cost differences that arise between the conventiona and condominia systems in terms of network expansion and socia intermediation. - The consumption differentia that exists between the conventiona and condominia systems as a resut of hygiene education activities. As far as possibe, the cost and consumption impacts of each of the four different components of the condominia approach engineering design, community participation, hygiene education and micro-credit wi be separatey identified. A costs and benefits wi be measured reative to the conventiona baseine. EL ALTO - BOLIVIA / PILOT PROJECT 19

22 WATER AND SANITATION PROGRAM Different anaytica perspectives The costs and benefits of the condominia system can be assessed from a variety of perspectives: that of the utiity, that of the water consumer, and finay the socia perspective. Utiity s perspective In order to simpify, it is assumed that the private operator is primariy concerned with profit maximization, or at the very east breaking even subject to the constraints paced by its contractua obigations. From this perspective, the key issue is the reative profitabiity of the conventiona and condominia systems, and hence the differing cost and revenue streams which they generate. In theory if the tariff structure perfecty refected the differing costs of service provision under the two systems the profit margin on water consumed from either type of connection woud be identica and the utiity woud be indifferent between the two technoogies, except in so far as the overa eve of water consumption was greater under one system than the other. The profit impact of seecting the condominia over the conventiona system woud then simpy be equa to the profit margin on each unit of water sod mutipied by the consumption differentia. Athough a definitive anaysis of the fu cost and tariff structure of the utiity ies beyond the scope of the present study, there are a number of reasons to think that current tariff structure may not be very cosey aigned with the structure of costs, even for conventiona systems. First, it is not cear how the current connection charges mandated by the concession contract reate to the actua cost of making connections The charges estabished in the concession contract of US$155 and US$180 for water and sewerage respectivey, fa substantiay short of the average connection costs of US$229 and US$276 reported by the utiity. However, this does not necessariy mean that the utiity is not covering the costs of network expansion since these may be partiay recovered from the genera use of service tariff. Second, the tariff structure does not incorporate a fixed charge to cover the administrative costs of meter reading and biing. According to AISA, these costs amount to approximatey US$1 per month. Given that the charge for the first cubic meters of water consumption is US$0.22, the impication is that househods which consume ess than five cubic meters per month do not even generate enough revenue to cover biing costs (Tabe 3-2). Third, the rising bock tariff structure is such that househods with ow eves of water consumption are not profitabe to serve. The exact breakeven eve of consumption depends on the true cost per cubic meter of water and sewerage services, a number that is not currenty known. However, there is itte doubt that it must be higher than the rate for the first band of domestic consumption which currenty stands at US$0.22. For exampe, the tariff study undertaken immediatey prior to privatization estimated an average cost for the water service of US$0.56 (Uzin, 1996). If this were the true average cost of water, a domestic customer woud have to consume more than 350 cubic meters of water per month before the revenue generated woud cover the cost of providing the service. Ceary, one woud expect the average cost of water (and hence the breakeven eve of water consumption) to have faen as a resut of the efficiency gains produced by the privatization process. However, the exact extent of these efficiency gains is not known at this point. Fourth, there is no separate tariff for the sewerage service. That is to say that there is a unified tariff that is supposed to cover the cost of both services, and which is paid by a customers regardess of whether they receive the sewerage service. The impication of this is that as the number of sewerage connections expands the utiity s costs increase without any concomitant growth in revenues. The average cost of water of US$0.56 reported above reates to the water service ony. If the (unknown) costs of the sewerage service were added in, the breakeven consumption eve woud increase further. 20

23 3. EVALUATION METHODOLOGY Finay, since condominia systems were not contempated under the origina concession contract their existence is not refected in the origina tariff structure. A differentiated connection charge has now been introduced so that customers connected under the condominia system enjoy a price reduction to refect the ower costs of the system. However, there is at present no differentiation in the use of service tariff paid by condominia customers who take responsibiity for maintaining their own network branches. For a of these reasons, the profit margins on the two types of connections are not necessariy the same, and may not even be positive. Hence, in order to assess the profit impact of the technoogy choice it is necessary to take costs and revenues fuy into account. Tabe 3-2 Tariff structure for AISA Consumption category (m 3 /month) Tariff Domestic Commercia Industria (US$/m 3 ) >301 >21 > as noted above, the consumer s perception of these costs may differ from that of the utiity. Given the difficuties with the current tariff structure identified above, the evauation wi be conducted using two parae scenarios. The first wi refect current reaity and using the existing tariff structure. The second wi refect an idea case and assume that tariffs perfecty refect costs. To summarize, the net benefits to the utiity (NBU) of the condominia system (D) are equa to the difference in the profit margin of the two technoogies where the profit margin is defined as the difference between the connection charge (CC) and the corresponding costs of connection (C) pus the profit margin on the additiona revenues (P) generated as a resut of higher consumption under the condominia system. The exact costs and revenues generated by the condominia system wi, of course, depend on the exact variant of the approach that is used. In particuar, community participation and hygiene education components wi tend to have higher associated costs but coud be expected to ead to higher service revenues. NBU D = (CC D C D ) (CC V C V ) + (P D P V ) Consumer s perspective The consumer s perspective differs from that of the utiity in three key respects. First, the consumer perceives the financia cost of materias and abor through the utiity s service charges. To the extent that tariff structures may not be perfecty aigned with cost structures, Second, the condominia approach is intensive in the use of a househod s time. Househods contribute their time and abor not ony for the construction of the condominia branches, but aso through their attendance of the workshops and community gatherings that form part of the condominia process. No financia cost is incurred since the utiity does not pay househods for their participation. However, uness househod members are competey ide, there is ikey to be an opportunity cost to their time which shoud be taken into account from an economic perspective. The question of what rate to use to vaue househod time is a difficut and controversia one. One approach is to vaue the househod s time according to the market cost of hiring somebody to perform the same activity. For exampe, to vaue the cost of a househod s time in constructing the condominia branches at the wage that is paid to a construction aborer; a rate of US$6 per day. However, this is probaby not the correct approach in this context. The reason is that the rate used shoud refect the vaue of a househod s time to the househod itsef. Hence, this rate woud ony be appropriate if the househod woud have been abe to use the time dedicated to the condominia EL ALTO - BOLIVIA / PILOT PROJECT 21

24 WATER AND SANITATION PROGRAM Figure 3-1 project to undertake activities generating US$6 per day. For a number of reasons, it seems unikey that this woud be the case. First, the minimum wage in Boivia is US$2.50 per day. Second, the average per capita earnings for workers iving in E Ato is US$4.14 per day 4. Third, the househod members participating in the condominia project woud not necessariy be the income earners. Anecdota evidence gathered from site visits suggests that women payed an important roe in the construction works. A number of women remarked that this was because their husbands woud have had to sacrifice wages if they had taken time off work. The opportunity cost of time for those who are not working, is much harder to gauge since it depends on the nature of the househod activities they perform. For the purposes of the anaysis, the minimum wage of US$2.50 per day wi be used as an indicative vaue of househod time. Third, the impact of increased water consumption is measured not by the additiona revenues to the utiity but rather by the increase in consumer surpus; which represents the wefare a househod obtains from consuming water over and above the price it pays. The difference between these two concepts can Difference between economic and financia benefits be iustrated graphicay in Figure 3-1, where the ine D V represents househod demand for water under the conventiona system. A condominia approach incorporating hygiene education or promotion of bathroom construction wi increase househod preferences for water consumption eading to higher consumption of water at a prices, and an outward shift of the demand curve to the higher ine D D. Within this framework, the hatched square area represents the increased financia revenues to the water utiity of which a sma component wi be the profit margin, whereas the soid trapezium shaped area represents the higher consumer surpus obtained by the househod. It is important to note that the change in consumer surpus may be arger or smaer than the increase in utiity revenues depending on the specific parameters of the demand curve. However, the two are unikey to be the same. To summarize, the net benefits to the consumer (NBC) of the condominia system (D) are equa to the associated savings in the connection charge (CC), minus the opportunity cost of time (OC) devoted to the condominia process, pus the greater consumer surpus (CS) resuting from higher eves of consumption. Once again, it is important to note that the extent of these benefits wi depend on the exact form of the condominia approach that is adopted. For exampe, when the engineering component aone is appied, the ony impact wi be the reduction in connection charge. When the community participation component is appied the connection charge differentia wi increase but the opportunity cost of time wi aso come into pay. Finay, when hygiene education activities are conducted, a change in the consumer surpus can additionay be expected. NBC D = (CC V CC D ) OC D + (CS D CS V ) Socia perspective The socia perspective is the most hoistic, taking costs and benefits to both utiity and customer into account. Thus the net benefits to society (NBS) 22

25 3. EVALUATION METHODOLOGY of the condominia system (D) are simpy the sum of the net benefits to the utiity and the consumer. The connection charges drop out of this equation since what is a cost to the consumer is a benefit to the utiity. Hence, from a socia perspective what matters is the underying investment cost differentia between the two systems (C), taking into account the opportunity cost of time absorbed by community participation (OC). To this must be added the additiona profits (R) accruing to the utiity from additiona water consumption and the additiona consumer surpus (CS) experienced by the consumer from the same. Comparison Finay, for the purposes of carity, the key differences between the evauation methodoogy from the three different perspectives are summarized in the tabe. Essentiay, the utiity s perspective considers financia costs, connection charges and subsequent profits from higher revenues. The consumer s perspective considers connection charges, the opportunity cost of time and subsequent increases in consumer surpus. Whie, the socia perspective considers financia and opportunity costs and both additiona profits and gains in consumer surpus. NBS D = (C V C D OC D ) + (R D R V ) + (CS D CS V ) Tabe 3-3 Comparison between anaytica perspectives Utiity Consumer Socia Cost of distribution networks Financia costs of materias and abor ü û ü Utiity connection and service charges ü ü û Opportunity cost of househod time in construction works û ü ü Benefits of water consumption Profits accruing to the utiity ü û ü Consumer surpus û ü ü EL ALTO - BOLIVIA / PILOT PROJECT 23

26 WATER AND SANITATION PROGRAM 4. Cost Differentias This section presents an estimate of the cost savings of condominia systems reative to conventiona ones, considering both financia costs and economic costs (such as the vaue of time). Separate treatment is given to network costs and the costs of socia intermediation. Under each heading, the financia and economic perspectives are separatey considered. Financia costs Network costs Investment costs The financia costs of the network expansion can be broken down into two components; abor and materias. First, even when the condominia branches are constructed by the community, abor must sti be contracted to undertake the upstream network extension and reinforcement. The costing of this component is straightforward because the figures can be taken directy from the contract invoices and the corresponding payment records in the utiity s database. Second, materias are needed to compete both the condominia branches and the upstream extensions. In both cases, materias are provided from the utiity s storage depot. Physica withdrawas from the depot are registered on a corresponding materias baance sheet. However, this documentation is not aways compete. Furthermore, the unit cost for each type of materia wi depend on when it was purchased by the utiity which may have been some time before the construction works. For both of these reasons, the vaues taken for the cost of materias are based argey on estimates that were undertaken by staff at the water utiity and members of the WSP consutant team. Since these estimates were independenty performed they are not identica, but are broady consistent in most cases. The average financia costs per water and sewerage connection for each piot site are presented in the figures beow (supporting tabes can be found in the Annex). For Phase One, where two sets of cost estimates are avaiabe the uncertainty range between the two estimates is shown as a fuzzy area in the figures. For Phase Two, ony the utiity s cost estimates are avaiabe and thus a singe point estimate is given. Unfortunatey, cost estimates are not avaiabe for a of the Phase Two sites, and in particuar for those neighborhoods where the condominia approach was appied without community participation. The reason is that these projects incorporated engthy 24

27 4. COST DIFFERENTIALS sewerage outfas whose cost coud not be fuy separated out in order to produce cost estimates comparabe to those presented for other sites 5. Finay, the Germán Busch piot was sti in progress at the time when the research for this study was carried out, hence the numbers given represent budgetary estimates as opposed to outturn costs. It is important to note that a figures are expressed as cost per connection, as opposed to cost per pot. In a the project neighborhoods the rate of connection to the new network was ess than 100%. This is argey due to the fact that on average ony 80% of the pots in any given neighborhood were occupied fu-time. The remaining 20% consist of undeveoped pots, or pots that are ony occupied for part of the year. The overa connection rate was 75%, however among fuy occupied pots the rate rises to 97%. Furthermore, the connection rates achieved by the EAPP for the water service (35%-80%) tend to be ower than for sewerage (70%-100%) refecting the fact that some househods aready had prior access to piped water. These uncon- Figure 4-1 Financia cost comparison for sewerage service Figure 4-2 Financia cost comparison for water service EL ALTO - BOLIVIA / PILOT PROJECT 25

28 WATER AND SANITATION PROGRAM nected ots ceary added to the cost of network expansion, but did not resut in an immediate connection. The costs per pot foow the same genera pattern as the costs per connection, but on average are 33% higher given that 75% of househods connect. The cost estimates presented beow are based on the assumption that the same proportion of househods connect to the network regardess of whether a conventiona or condominia system is provided. However, there is some evidence that connection rates may be higher in condominia than conventiona systems, perhaps owing to the greater invovement and ownership experienced by the community. For exampe, in 3 de Mayo (a contro neighborhood recenty connected to the conventiona sewerage network) ony 66% of househods made a connection (Cannei, 2001), compared to an average of 75% across the condominia neighborhoods. To the extent that such a difference exists, the costs per connection given for conventiona sewerage woud tend to be understated and hence the cost savings of the condominia approach woud aso be understated. It was not straightforward to come-up with unit cost figures for conventiona sewerage against which to compare the costs of the condominia systems. There has been ony one conventiona sewerage project in E Ato, in a neighborhood known as Brasi. This project resuted in a cost of US$240 per connection. However, it was based on the use of cement pipes which are consideraby ess expensive than the PVC pipes used in the EAPP sites and hence as a point of comparison it is too ow. Moreover, the fact that it is a singe site cost estimate as opposed to an average of severa sites does not make it an idea reference point, since it may be distorted by site-specific idiosyncrasies. AISA has aso done five conventiona sewerage projects in ow income neighborhoods in La Paz, that perch on the steep ciffs or aderas surrounding the city. The average cost per conventiona sewerage connection for these neighborhoods has been in the range US$265 to US$410, with an average of US$ However, the topographica conditions of the aderas are very different to those found in E Ato and hence these sites do not provide an appropriate point of comparison with the EAPP sites. In concusion, historic experience with conventiona sewerage projects in La Paz and E Ato does not appear to provide the necessary cost comparator for the EAPP sites. A that can reiaby be deduced from this experience is that the cost per connection for conventiona sewerage in E Ato woud probaby ie somewhere in the range US$240 to US$310. Therefore the approach adopted was to take as a reference point the average estimated connection cost of conventiona sewerage for those Phase One neighborhoods where detaied cost estimates had been undertaken. This ed to an estimated cost per conventiona sewerage connection of US$276, which in fact turns out to be consistent with the range defined above. In the case of water, conventiona services have been provided in three neighborhoods in E Ato, with connection costs ranging from US$140 to US$216, and an overa average of US$194. This average is simiar, but sighty ower than, the average of US$229 obtained from the Phase One EAPP neighborhoods where conventiona water service cost estimates were made. For the sake of consistency with the method used for the sewerage service, this second number is used in making the cost comparisons reported in the figures beow. The main resuts are as foows. For sewerage services, the average cost per connection with community participation is US$142, with a range from US$107 to US$176 depending on the neighborhood. The range of uncertainty in Phase One cost estimates is sma in reation with the variations in unit costs across project sites. Overa, this represents an average cost saving of US$134 (or 48%) on the corresponding cost of a conventiona sewerage connection which stands at US$276. For water services, the genera pattern of resuts is very simiar. The average cost per connection 26

29 4. COST DIFFERENTIALS is US$112, with a range from US$93 to US$149 depending on the neighborhood. Once again, the uncertainty ranges in Phase One estimates are sma in reation to the unit cost variations across piot neighborhoods. These resuts suggest an average reduction of US$117, equivaent to 51% of the cost of a conventiona water connection which stands at US$229. Figure 4-3 Breakdown of cost savings for sewerage service In summary, the overa pattern of resuts is that the condominia approach approximatey haves the financia costs of network expansion for both water and sewerage systems when the community participates in the construction process. These savings are broady consistent with those reported for condominia systems in Brazi. For exampe, Bakaian et a. (1988) find cost savings of around 40% for sewerage in Brazi. Whie, a recent survey of 10 years of experience with condominia sewers in Brasiia, concuded that cost savings had been of the order of 40%-50% (Luduvice et a., 2001). Larger savings have been reported in some cases; such as the city of Petroina where reductions of 60%-80% have been reported (Watson, 1994). Figure 4-4 Breakdown of cost savings for water service Breakdown of cost savings The cost savings observed with the condominia mode potentiay come from three different sources. First, the savings in the cost of materias as a resut of the shorter ength and narrower diameter of the pipes. Second, the savings in abor effort resuting from the shorter and shaower trenches that can be used in the condominia case. Third, the savings in abor costs that arise from community participation. The first two of these are avaiabe when the condominia approach is appied without community participation, whereas the third is directy attributabe to the community s vounteer abor. By breaking down the cost savings experienced in the EAPP neighborhoods, it is possibe to quantify the reative importance of these three factors (Figures 4-3 and 4-4). The main resuts are as foows. Materias savings are reativey modest, accounting for ony 10% of the overa cost reduction. There are two expanations for this. First, materias account for ess than haf of the tota costs of conventiona systems (36% for sewerage and 47% for water). Second, the physica savings made in the ength and diameter of pipes are comparativey sma; of the order of 10%- 20%. Conversey, abor savings account for around 90% of the tota cost reduction. As before, this refects the fact that abor represents a higher share of the tota costs of conventiona systems (64% for sewerage and 53% for water). Moreover, the physica savings in the voume of soi to be excavated are substantia, at 45% for water and 75% for sewerage. EL ALTO - BOLIVIA / PILOT PROJECT 27

30 WATER AND SANITATION PROGRAM The breakdown of abor savings between abor effort and abor cost varies substantiay across services. Labor cost savings dominate for the sewerage service, whie abor effort savings dominate for the water service. From these resuts, it is possibe to infer that the condominia design appied without community participation woud have ed to cost savings of 24% for the sewerage service and 40% for the water service. These resuts are consistent with recent estimates reported by Aguas de Iimani for a new project site in E Ato caed Maasia. The Maasia cost estimates suggest savings of 22% for the condominia approach without community participation, as against 42% with community participation. They are aso consistent with evidence from CAESB, the water utiity of Brasiia, which reports cost savings of the order of 20% for sewerage systems without community participation (Neder, 2001). Impact of connection rates The cost savings estimated above are based on the impicit assumption that the proportion of househods connecting to a sewerage network wi be the same regardess of whether a conventiona or condominia approach is adopted. There is some evidence to suggest that this may not be the case. Neighborhoods that formed part of the EAPP achieved an average sewerage connection rate of 75%, whereas in a contro neighborhood (3 de Mayo) the sewerage connection rate was 66%. The fact that there was ony a singe contro neighborhood makes it difficut to know how much weight to attach to this resut. However, if the effect of this differentia in connection rates is factored into the cacuation of cost savings, the overa cost saving from condominia sewerage increases from around 40% to 45%. Operating and maintenance costs In theory, the condominia system with community participation shoud aso have the effect of reducing the operating and maintenance costs of the sewerage network. However, in practice this is very difficut to quantify. One reason for this is simpy the difficuty of disaggregating company cost data to the eve required to identify differences in maintenance costs over reativey sma segments of the network. A second reason is the nature of maintenance activities. Particuary in the case of sewerage, maintenance on the finer branches of the distribution network tends to be reactive and consequenty infrequent. It woud therefore be necessary to track highy disaggregated cost data over a fairy ong period of time in order to be abe to identify such differences. In the case of E Ato, both probems are pertinent. On the one hand, historicay there has not been separate accounting between water and sewerage systems in La Paz and E Ato, et aone disaggregated information for the sewerage service. AISA is required to introduce separate water and sewerage accounting under the terms of the concession, however more disaggregated information is not yet avaiabe. On the other hand, the very recent arriva of the condominia system in E Ato means that insufficient time has accumuated to be abe to estabish the ong term average annua maintenance costs. Unfortunatey, the iterature on the Braziian experience is not very concusive on this subject. Bakaian et a. (1988) were unabe to obtain any data to document the maintenance cost differentia. Watson (1994) reports some maintenance data for the city of Petroina, which shows that the maintenance cost saving to the utiity when customers maintain their own sewers is ess than 10%. Watson aso reports that whie tariff reductions of around 50% are typica for condominia neighborhoods in Brazi, this is an ad hoc reduction that is not based on a scientific comparison of maintenance costs. For the case of Brasiia, Luduvice et a. (2001) report that the number of maintenance visits per kiometer of sewer per month is simiar between condominia and conventiona systems, but that the cost per maintenance visit is substantiay higher for conventiona systems. Overa, they find that maintenance costs are around 70% ower for condominia systems. The current practice in Brasiia is for condominia customers to receive a 28

31 4. COST DIFFERENTIALS discount of around 40% on their sewerage bis reative to conventiona customers. Tabe 4-1 Contracting of socia work Costs of socia intermediation The preceding discussion considered ony the cost of network expansion. However, the condominia system aso entais significant costs in socia intermediation, which need to be taken into account in estimating the fu financia cost. The socia work required for the condominia method was contracted out during Phases One and Two of the project (Tabe 4-1). The first contract, reating to the neighborhoods of Huayna Potosí and Via Ingenio, was awarded to an NGO: CIEC. Owing to the high costs of the services provided by CIEC, the remainder of the Phase One neighborhoods were contracted to ESUE, a private company formed by a number of socia workers previousy empoyed by CIEC who had participated in the origina piots. ESUE was aso awarded the contract for the Phase Two neighborhood, E Ingenio. However, in the case of Germán Busch, AISA preferred to enter into individua contracts with four socia workers who were former empoyees of ESUE, and hence had experience from earier piot neighborhoods. As regards the future, AISA has decided that from henceforth it wi use its own in-house staff for the socia intermediation activities. The costs of socia intermediation can be divided into two categories. On the one hand, there is the cost of the contracted socia workers. On the other hand, the cost of the WSP (Phase One) or AISA (Phase Two onwards) staff responsibe for supervising the work of the contractors must aso be taken into account. Tabe 4-2 summarizes the evidence regarding the costs of the contracted socia workers. The comparison between neighborhoods is compicated by various factors. First, there is a earning effect which impies that the first neighborhood to be covered (Huayna Potosí) has a very high number of man-days of socia intermediation per connection (11.9). Second, the man-days of socia intermediation Contractor Vaue Projets Covered CIEC US$65,072 Huayna Potosí Via Ingenio D-II ESUE US$71,000 Caja Ferroviaria San Juan de Río Seco Oro Negro Jichusirca ESUE US$21,000 E Ingenio Individua US$7,897 Germán Busch Tabe 4-2 Costs of socia intermediation Man-days per Cost per Cost per connection man-day connection US$ US$ Phase One Water ony - Jichusirca Sewerage ony - Huayna Potosí Oro Negro Water and sewerage - Via Ingenio D-II Caja Ferroviaria San Juan de Río Seco Phase Two Water and sewerage - E Ingenio U-IV Germán Busch per connection aso vary according to whether water, sewerage, or both services were being provided to the community. Third, the cost per man-day in the first contract (awarded to CIEC for Huayna Potosí and Via Ingenio) at US$28 per man-day is substantiay higher than the cost of US$19 per man-day found in subsequent contracts. Fourth, the socia intermediation conducted under Phase One had a strong hygiene education component. This activity was substantiay scaed down in Phase Two, due to the fact that the Boivian government began to provide subsidies to househods in E Ato to construct their own bathrooms. EL ALTO - BOLIVIA / PILOT PROJECT 29

32 WATER AND SANITATION PROGRAM Taking a of these factors into consideration, the foowing concusions appear to emerge from Tabe 4-2. Under Phase One, the average number of man-days per connection after the earning experience of Huayna Potosí, is 2.3. Taking the ower rate of US$19 as the cost per man-day, this works out at US$44 per househod connected to both water and sewerage services. The costs for connections to a singe service are substantiay ower. Namey, US$13 per connection for water (based on Jichusirca) and US$26 per connection for sewerage (based on Oro Negro). The cost per sewerage connection is about twice as high as per water connection. This cost differentia is argey attributabe to the fact that househods receiving sewerage received additiona support from socia workers in order to hep them to construct bathrooms. From this it may be inferred that the bathroom support component cost around US$13 per househod. The costs of socia intermediation under Phase Two, at an average US$15 per househod connected to both water and sewerage services, are amost a third of the costs registered under Phase One. These differences are due to number of factors, incuding earning effects, a more costconscious approach to supervision by AISA, and a substantia reduction in the scope of the socia work due to the excusion of the bathroom support component. For the purposes of this evauation exercise the US$15 cost wi be used because it embodies the fu earning effects accumuated during the various piots. This vaue averaged across the two services works out at about US$8 per househod per service. Given the absence of Phase Two experience with hygiene education, the Phase One estimate of US$13 per househod wi have to be used, even though it seems ikey that this cost woud have reduced somewhat in Phase Two as efficiency eves improved. It is important to note that these costs of socia intervention do not take into account the costs of WSP consutants invoved in the EAPP, which came to just under US$720,000 over the three year period of the project: These costs are evidenty high in reation to the sum of around US$1 miion invested in the physica networks, and particuary in reation to the US$0.7 miion that in aggregate was saved as a resut of providing condominia rather than conventiona sewers. However, to attempt to aocate these costs to individua connections made in the project woud be miseading. Essentiay, these represent an investment in the transfer of the condominia know-how from Brazi to Boivia, entaiing the adaptation of the approach to the oca context, the training of oca professionas in the impementation of the methodoogy, and a number of monitoring and evauation and dissemination activities. As can be seen the costs of this transfer are substantia, but they shoud be amortized across a future condominia connections in Boivia and not simpy those carried out in the EAPP. It is estimated that of the tota expenditure under the WSP program 55% corresponds to the adaptation process, 25% to the training process, and 20% to other activities such as monitoring, evauation and dissemination among others 7. Summary of financia costs Tabe 4-3 beow summarizes the financia costs of conventiona versus condominia water and sewerage services. Condominia systems ead to savings in network expansion costs of around 50% for both water and sewerage. The costs of the socia intermediation required to engage community participation prove to be modest in reation to the associated cost savings. The overa cost advantage of condominia systems is reduced to around 45% when socia intermediation is taken into account. Economic costs Network costs The financia costs reported above are estimated from the perspective of the water utiity. In order to move from financia to economic costs it is necessary to shift to the customers perspective. This entais two changes. 30

33 4. COST DIFFERENTIALS First, it is necessary to incorporate the opportunity cost of the time devoted to the construction of the condominia networks. Second, the monetary cost of the connection to the consumer is the connection charge as opposed to the costs reported in the preceding section. Opportunity cost of time In order to vaue the opportunity cost of time for community participation, two different approaches wi be used. The first is an activity based approach, which was appied to the Phase One sites by the WSP team. The idea is to ist the tasks that were performed by the community at each project site and then to scae down the commercia unit cost of these activities by the ratio between the community s vaue of time (US$2.50 per day as derived above) and the daiy wage of a construction worker (US$6 per day as noted above). The resuts of this approach suggest that the vaue of community time is about US$12 for a water connection and US$16 for a sewerage connection (Tabe 4-4). The aternative approach is to estimate, based on fied observations, the tota number of man-days that each beneficiary was putting into the construction of the condominia networks, to which can be appied the daiy vaue of time. The resuts are US$15 per water connection and US$17 per sewerage connection (Tabe 4-5). Given that the two approaches give reasonaby consistent vaues, average vaues of US$14 per water connection and US$17 per sewerage connection, wi be adopted in the subsequent anaysis. Over and above the time contributed for construction of condominia networks, it is estimated that routine maintenance of condominia branches undertaken by househods takes approximatey 8 hours per month, equivaent Tabe 4-3 Water Sewerage Conventiona Condominia Conventiona Condominia Network expansion Socia intermediation Tota Tabe 4-4 Water Sewerage Huayna Potosí 11 Via Ingenio (D-II) Caja Ferroviaria 8 12 San Juan de Río Seco 6 21 Oro Negro 21 Jichusirca 13 Average Tabe 4-5 Summary of financia costs (US$) Activity based costing (US$) Time based costing Water Sewerage Days US$ Days US$ Preparatory work Digging of trenches Laying of pipes Laying of septic chambers 0.70 Refi of trenches Tota to a vaue of around US$2 (Programa de Agua y Saneamiento, 2001). Connection charges Connection charges are reported in Tabe 4-6. Foowing the experience of Phase One, AISA has introduced a standard differentiated connection charge for condominia water and sewerage services. These currenty stand at US$100 for each of the two services, athough in practice the water connection charge incudes an additiona US$36 to cover the cost of the water meter. For the consumer, this represents a saving of US$19 (or 16%) in the case of water and US$80 (or 44%) in EL ALTO - BOLIVIA / PILOT PROJECT 31

34 WATER AND SANITATION PROGRAM Tabe 4-6 the case of sewerage. However, there is no charge reduction for condominia connections undertaken without community participation. Finay, since these charges are nonetheess high in reation to househod income, AISA aows househods to pay the connection charge in monthy instaments over a two year period. The interest rate appied to these monthy instaments is 13% per annum. This is approximatey haf of the minimum interest rate charged by oca microfinance institutions. Water Sewerage (US$) (US$) Origina Concession Contract 119* 180 Currenty Condominia with community participation 100* 100 Condominia without community participation 119* 180 Conventiona 119* 180 *For purposes of comparison the US$36 cost of the meter has been excuded. Tabe 4-7 Connection charges Opportunity cost of socia intermediation Duration Attendance Vaue of (hours) (hours) Time (US$) Workshops Presentation of project Presentation of network design Training for construction work Reguations of condominia Maintanance techniques Hygiene habits Construction of bathrooms House Visits Construction of network Construction of bathrooms Tota time Construction of network Construction of bathrooms Overa Costs of socia intermediation Just as the time devoted by the community to the construction of condominia networks has an opportunity cost, so does the time devoted by the community to participating in the activities reating to the socia intermediation process. Tabe 4-7 ists a of the different workshops and house visits that are invoved in the impementation of the condominia approach together with the estimated duration of each. The figures show that the socia intermediation associated with community participation in the construction of condominia networks, can absorb 24 hours per househod, over the 3-6 month duration of the project. The incusion of hygiene education and bathroom support activities add a further 10 hours to the participation of each househod. According to househod survey data coected by the WSP team, each event was actuay attended by around 75% of participating househods. The opportunity cost of this time is vaued using the rate of US$2.50 per day (or US$0.25 per hour) estabished above. On this basis, the overa opportunity cost of socia intermediation comes cose to US$5 for participation associated with the construction of the network, and US$2 for activities reating to hygiene education and bathroom support. Arguaby the opportunity cost of time attaching to socia intermediation activities is ower than that reating to the construction of condominia branches. The reason is that workshops and house visits were typicay conducted outside of norma working hours, that is to say in the evenings and weekends. Costs of intra-househod investments In order to be abe to benefit from a sewerage connection, househods need to construct a bathroom of their own. The bathrooms constructed by househods in the project area are typicay buit from scratch, and invove the construction of an independent out-house containing a toiet, hand basin and eectric shower. The eve of sophistication of these faciities varies consideraby across househods. However, the average cost of constructing a 32

35 4. COST DIFFERENTIALS bathroom is estimated to be of the order of US$440 (Tabe 4-8). This overa cost can be broken down into US$400 for materias and about US$40 for abor, which it is assumed wi be suppied by the househod itsef. This represents a very substantia cost for househods in the project area. Indeed, it is about twice as high as the sewerage connection charge and is equivaent to about one month of average househod income. In order to ensure that this did not create a financia barrier for househods to benefit from the sewerage connection, the project incuded a micro-credit ine to provide oans for bathroom construction. A second tier bank, FUNDAPRO, offered to provide up to US$500,000 of capita for the microcredit scheme. Foowing a bidding process, the retaiing of the oans was contracted to Caja Los Andes (CLA), a micro-credit institution with an estabished brand name in the project area. This was the first time that CLA had offered oans to househods specificay for utiity services. CLA was wiing to suppy oans to househods with a good credit history as ong as certain basic financia conditions were met. In particuar, the vaue of the oan shoud not exceed 70% of the househod s assets. As a specia inducement to househods, CLA was wiing to consider owner occupied housing as coatera; a departure from its standard procedure of considering ony movabe assets (such as teevisions, furniture, etc.). In addition, the monthy repayment shoud not exceed 70% of the vaue of the househod s surpus income defined as income minus an estimated iving cost of the order of US$20-US$30 per person per month to cover food, basic services, transport, education, heath and contingencies. As a further inducement to househods, CLA offered terms of 1.92% interest per month on oans made under the project, which represents a significant discount on their standard rates of 2.50% per month. The impications for monthy repayment charges on typica terms of up to 36 months are iustrated in Tabe 4-9. In spite of these preferentia terms, the eve of participation from househods was ower than expected. Some 25% of Phase One househods Tabe 4-8 Average Instaation Costs (US$) Materias Labor* Tota Phase One Huayna Potosí Via Ingenio (D-II) Caja Ferroviaria San Juan de Río Seco Oro Negro Jichusirca Overa Source: estimates of WSP team members based on non-random samping of househods * Based on 16 man-days of abor vaued at US$2.50 per day. Tabe 4-9 Average cost of bathroom construction Reference tabe for cost of debt servicing Principa Monty repayment (US$) for a term of (US$) months months months months months Source: Caja Los Andes, La Paz, Boivia. appied for credit, of which 75% met with the CLA creditworthiness conditions. However, this overa average conceas significant variation between the two Phase One neighborhoods, with more than 50% of househods in Huayna Potosí appying for a oan, as against fewer than 20% of househods in Via Ingenio (Tabe 4-10). Overa, ony US$43,000 of the origina capita coud be paced, ess than 10% of the tota originay earmarked for the project. Owing to reativey ow eves of participation in the Phase One neighborhoods, the micro-credit ine was suspended from Phase Two onwards. The average borrowing househod had a monthy income of US$562 (see tabe and figure). This vaue is we above the poverty ine and extreme poverty ines reported in Tabe 2-1 above. As reported there, about 60% of the popuation of E Ato ive in poverty and 40% in extreme poverty. By contrast, ess than 25% of borrowers ive beow the poverty ine and ess than 15% EL ALTO - BOLIVIA / PILOT PROJECT 33

36 WATER AND SANITATION PROGRAM beow the extreme poverty ine. The absence of income data for non-borrowing househods makes it difficut to draw rigorous concusions. Nevertheess, these comparisons in combination with the other economic characteristics of participating househods (high rates of home ownership and significant surpuses of income over out-goings), suggest that take-up of oans was concentrated disproportionatey among nonpoor househods. This finding is consistent with a recent study of micro-credit institutions in Boivia, which notes that their cients tend to be peope iving cose to the poverty ine as opposed to those iving beow it (Navajas et a., 2000). The same study finds that among Boivian micro-credit institutions, CLA reaches a ower proportion of the poorest than some of its competitors. Participating househods on average borrowed a principa of US$407, with a term of 32 months. As a resut, they incurred monthy repayment charges of US$14.21; equivaent to 6.2% of income or 35.3% of surpus income (after netting out basic iving costs). The principa to asset ratio for the average borrowing househod was 21.9%. About 86% of househods were abe to take advantage of the possibiity of using their house as coatera. The average vaue of the dweings occupied by borrowing househods was estimated by CLA to be of the order of US$4,291. However, the vaue of the coatera that coud be raised by non-owner occupiers was substantiay ower at US$ A number of reasons have been advanced to expain the ower than expected take-up of the micro-credit ine. The first possibe expanation reates to the way in which the oans were administered. CLA deegated the marketing of the credit ine to the project socia workers, who were evidenty not Tabe 4-10 Take-up of oans by neighborhood Huayna Potosí Via Ingenio Tota # % # % # % Obtained credit Refused credit Managed without Tota Source: Caja Los Andes, La Paz, Boivia Tabe 4-11 Characteristics of borrowing househods Income Out-goings Surpus Home Vaue of Vaue of other US$ US$ ratio ownership home coatera US$ US$ Average % 86.0% 4, Quarties - First % 1, Second % 2, Third % 6, Range - Minimum % 1, Maximum 3, , % 8, Source: Caja Los Andes, La Paz, Boivia 34

37 4. COST DIFFERENTIALS speciaists in this fied. CLA participation was imited to sending aong its own representatives about once per week to process credit appications. The company hypothesizes that higher penetration rates might have been achieved had a dedicated team of company representatives been assigned to the project 9. The second possibe expanation reates to the preferences of the oca peope. According to the experience of CLA 10, Aymara peope prefer to use any weath they may accumuate to make investments in rea estate or to finance community festivities. They are ess incined to spend money on improving persona comfort eves, for exampe by constructing a bathroom. Furthermore, given that a bathroom does not yied a direct financia return peope may be unwiing to go into debt in order to finance it. The third possibe expanation reates to the existence of aternative finance mechanisms. The Aymara peope often prefer to make use of their own traditiona financing mechanisms, such as the pasanaco, because these do not entai the expicit payment of interest. The pasanaco is an arrangement whereby a group of peope typicay reatives or friends pay a reguar amount into a common fund; every so many months the entire vaue of the fund fas to one of them for their discretionary use. The fourth possibe expanation reates to the wider economic and financia conditions in the country at the time of the project. Boivia was entering an economic downturn at around the time the micro-credit scheme was being advanced, which brought about a fa in househod incomes. Furthermore, the country had aso just come through a consumer credit crisis caused by aggressive marketing of oans by a number of institutions. Both factors may have made consumers more wary of participating in the credit scheme. It is not possibe to say which of these four factors was the most important factor behind the reativey ow rate of take-up for the oans. Summary of economic costs Tabe 4-13 beow summarizes the economic costs of conventiona versus condominia water and sewerage services. Two separate sets of resuts are reported. The first is based on connection charges as they currenty stand, and hence refects the Figure 4-5 Income distribution of househods participating in micro-credit scheme EL ALTO - BOLIVIA / PILOT PROJECT 35

38 WATER AND SANITATION PROGRAM rea experience of househods. The second is based on connection charges as they woud be if they truy refected the financia costs estimated above. The first set of resuts show that, from the consumer s perspective, there is currenty no rea advantage to taking a condominia water connection. The reason is that the reduction in the connection charge (from US$119 to US$100) is not arge enough to offset the opportunity cost of time invoved in participation which comes to US$19. On the sewerage side, however, there is a cost advantage of about US$60, equivaent to a reduction of over 30%. The second set of resuts show that, when charges are adjusted to cost-refective eves, condominia water connections do ead to a US$90 saving equivaent to a reduction of around 40%. The saving for condominia sewerage connections is over US$100, again equivaent to a reduction of around 40%. Finay, at an estimated tota of US$443, the time and money costs of bathroom construction are high in reation to those reating to the expansion of the network. When these are factored in, the fu cost of the sewerage service under the hygiene education variant rises to US$628. Tabe 4-12 Characteristics of credits granted Principa Term Repayment Principa to Repayment Repayment (US$) (months) (US$) asset ratio to income to surpus (%) ratio (%) ratio (%) Average Quarties - First Second Third Range - Minimum Maximum Source: Caja Los Andes, La Paz, Boivia Tabe 4-13 Summary of economic costs (US$) Water Sewerage Conventiona Condominia Conventiona Condominia Actua charges - Connection charge 119* 100* Network expansion Socia intermediation Tota Cost-refective charges - Connection charge Network expansion Socia intermediation Tota *Excudes US$36 cost of meter. 36

39

40 WATER AND SANITATION PROGRAM 5. Consumption Differentias As we as eading to savings in the costs of network expansion and maintenance, the condominia system may aso ead to higher eves of water consumption, thereby increasing the revenues of the water utiity and the wefare of the water consumer. The consumption effect is thought to come from the hygiene education component of the socia intermediation activity, which supported househods in the construction of a bathroom and reated househod instaations (such as kitchen sinks). This section first tries to isoate the consumption differentia attributabe to the condominia approach, and then goes on to vaue this differentia both from the financia and the economic perspectives. Estimating the consumption differentia The anaysis of consumption is based on househod survey data coected by the WSP team on the compete sampe of beneficiary househods from Phase One, pus a contro neighborhood in E Ato (3 de Mayo) which receives conventiona as opposed to condominia water and sewerage services. The WSP survey coected information on a wide range of househod characteristics, incuding of particuar interest for this purpose water-reated instaations within the househod. On the basis of the utiity s customer code, it was subsequenty possibe to match individua househods in the WSP database with those in AISA s biing database, so that a series of monthy water consumption readings coud aso be incorporated into the dataset. There are two stages to the process of determining the consumption differentia attributabe to the condominia approach. First, it is necessary to estabish to what extent the incusion of a hygiene education component in the condominia approach increased the rate of construction of bathrooms and reated househod instaations. Second, it is necessary to estabish to what extent these bathrooms and reated instaations increased househod water consumption on average. Impact on househod instaations Tabe 5-1 compares the prevaence of waterreated househod instaations in the EAPP neighborhoods, with that in the contro neighborhood (3 de Mayo). Athough the coverage of domestic instaations varies substantiay across condominia neighborhoods, by a range of up to 30%, a condominia neighborhoods have higher rates of coverage of 38

41 5. CONSUMPTION DIFFERENTIALS water-using faciities than the househods in 3 de Mayo (Figure 5-1). Thus, bathroom construction is on average 35% higher in the condominia neighborhoods than in 3 de Mayo. Moreover, whie the ony instaation that househods in 3 de Mayo have is a water coset, a significant proportion of househods in the condominia neighborhoods have instaed other water-reated faciities such as hand basins, showers, kitchen sinks and aundry sinks. The regression mode reported in Tabe 5-2 aims to uncover some of the factors that might expain the variation in bathroom construction rates across neighborhoods within the condominia piot. Curiousy, the avaiabiity of a micro-credit faciity does not appear to have had a significant effect on the rate of bathroom construction. However, the foowing three factors seem to have been important. First, when a house is continuay, as opposed to occasionay, inhabited the probabiity of constructing a bathroom increases by Second, providing water and sewerage service simutaneousy raises the probabiity of bathroom construction by 0.33, as against Tabe 5-1 Coverage of water-reated instaations in househods with sewerage Condominia Neighborhoods Averages Huayna Via Caja San Oro Conventiona Potosí Ingenio Ferroviaria Juan Negro Condominia (3 de Mayo) Differentia Bathroom Instaations - Water coset Hand basin Shower Kitchen sink Laundry sink Tabe 5-2 Resuts of regression mode for probabiity of bathroom construction* Coefficient Standard T- Significant Impact on error statistic at 95% probabiity eve? of bathroom construction Househod size No Aways inhabited Yes Combined water and sewerage Yes Intensity of socia work Yes Avaiabiity of microcredit No Constant No Regression parameters Observations 1292 Log ikeihood function Adjusted R squared 0.15 Chi-squared statistic (5) *Regression is performed ony on those househods who have sewerage service. 39

42 WATER AND SANITATION PROGRAM Figure 5-1 Prevaence of water reated instaations in househods with sewerage connections neighborhoods that had aready had the water service for some time before the sewerage service was extended. Third, the intensity of socia work (measured in terms of days per connection) aso had a significant effect. An extra day of socia work per connection raises the probabiity of bathroom construction by In summary, househods in condominia neighborhoods raised their chances of constructing a bathroom on average by 35%. The intensity of socia work appears to have been an important reason for this. However, it is not the ony one. Factors such as the joint provision of water and sewerage services may have had a arger roe to pay. Impact on water consumption The househod database indicates that monthy water consumption in E Ato is very ow by internationa standards, with the average househod taking 5.91 cubic meters per month, equivaent to just under 40 iters per capita per day. Indeed, amost 50% of househods consume ess than 5 cubic meters per month (Figure 5-2). To put this in context, an often used internationa benchmark for subsistence consumption is 120 iters per capita per day (equivaent to 18 cubic meters per month). However, fewer than 3% of househods in E Ato consume above this eve. This section considers the extent to which househods with bathrooms registered higher water consumption than househods without bathrooms. On the basis of the avaiabe data, there are two possibe ways of capturing this consumption differentia. Time series approach: In the two neighborhoods (Huayna Potosí and Oro Negro) that had had the water service for some time before the sewerage service, it is possibe to compare how water consumption for the same househod changes when sewerage arrives and in particuar when a bathroom is instaed. Cross-sectiona approach: Esewhere, water and sewerage services were provided simutaneousy, and hence no before and after comparisons are possibe. However, since the sampe contains a significant number of househods which have water services but ack sewerage services, it is possibe to make crosssectiona comparisons between the two. The time-series anaysis is reported in Tabe 5-3. In order to minimize as far as possibe the distortion 40

43 5. CONSUMPTION DIFFERENTIALS created by short-run seasona fuctuations, the water consumption variabe used for the purposes of comparison is a six month average 11. That is to say that the average monthy consumption in the six months immediatey after the arriva of the sewerage service, is compared with the average monthy consumption for the same six month period during the year preceding the arriva of the sewerage service. In order to isoate the effect of bathroom construction, separate comparisons are undertaken for househods with and without bathrooms. The resuts indicate that in both neighborhoods as might be expected, househods that did not construct bathrooms saw no significant change in their water consumption foowing the arriva of the sewerage service 12. However, consumption did increase significanty in househods that constructed their own bathrooms 13. In Huayna Potosí, househods who constructed bathrooms were consuming on average two cubic meters per month more than they did beforehand, a growth of over 30%. In Oro Negro, the differentia is somewhat smaer at 0.6 cubic meters per month, equivaent to growth of 8%. Figure 5-2 Consumption histogram Turning to cross-sectiona evidence, Tabe 5-4 iustrates the consumption differentias between househods with different instaations. Consumption is measured as the monthy average over the period Apri to August 2000, which was the ongest period for which consumption data was consistenty avaiabe in a of the piot neighborhoods. The pattern of resuts is quite consistent with that found in the time-series anaysis. The difference in water consumption between househods without sewerage and those Tabe 5-3 Average consumption before and after sewerage Observations Mean Standard 95% confidence interva error Lower bound Upper bound Huayna Potosí (a) Sewerage connections without bathrooms Before After Difference (b) Sewerage connections with bathrooms Before Afer Difference Oro Negro (a) Sewerage connections without bathrooms Before After Difference (b) Sewerage connections with bathrooms Before After Difference

44 WATER AND SANITATION PROGRAM that have sewerage but have not constructed bathrooms is about one cubic meter per month; but is ony just statisticay significant 14. Househods with bathrooms, consume on average 1.9 cubic meters per month more than househods with sewerage but without bathrooms. The difference is statisticay significant 15. The remainder of Tabe 5-4, expores the impact of specific sanitary instaations on monthy water consumption. Thus, for exampe, househods with a water coset consume on average two cubic meters per month more than househods without one. In the case of handbasins, showers, kitchen sinks and aundry sinks, the differentia is at east three cubic meters per month. However, these simpe averages are miseading because there is a high correation between househods that have water cosets and househods that have showers and other instaations. In order to isoate the contribution of specific sanitary instaations, mutipe regression techniques are required that contro simutaneousy for the presence of a of these devices. To this end, Tabe 5-5 presents the resuts of a inear regression mode that succeeds in expaining 32% of the variation in average monthy water consumption across househods in terms of a range of expanatory variabes. There are ony two types of variabes that appear to have a significant effect. The first type are variabes reating to habitation and househod size. In particuar, it is found that each additiona househod member adds on average about haf a cubic meter per month to the tota consumption of the househod. The second type are variabes reating to househod instaations. Househods with a shower on average consume one cubic meter a month over and above househods that do not have a shower. A simiar differentia is found for the presence of a aundry sink, whie a kitchen sink adds about one and a haf cubic meters to monthy househod consumption. As interesting as the variabes that prove to be significant in this mode are those that do not turn out to be significant. For exampe, it is striking that the presence of a water coset does not appear to add significanty to househod water consumption. Nonetheess, this resut accords with discussions hed with househods in the fied, who stated that they thought it extravagant to fush the toiet with cean water and therefore tended to recyce water from other uses for this purpose. The fact that the presence of sewerage does not have any significant effect on househod water consumption merey confirms the resut reported above, that what matters is not the sewer itsef but the number of househod instaations to which it is connected. A simiar argument expains the absence of any statisticay significant effect for the presence of a bathroom. The bathroom itsef is no more than a room, what matters for water consumption are the faciities that it contains. Impact on waste water behavior Househods receiving hygiene education were aso found to significanty reduce insanitary practices. The percentage throwing out used water into the streets fe from 77% before the project to 58% thereafter, and the percentage recycing water within the home fe from 36% to 25% (Cannei, 2001). Unfortunatey, it is not straightforward to put a monetary vaue on these behaviora changes. However, they are unquestionaby important in terms of bringing about improvements in environmenta quaity and human heath. Vauing the consumption differentia The baance of evidence reported above suggests that in neighborhoods where the hygiene education component of the condominia approach were appied, the average impact on consumption was 0.7 cubic meters per month 16, or an average increase of about 10% 17. In this section, the vaue of this consumption increment is estimated both from the utiity s and the consumer s perspective. Financia perspective As outined in Section 3, under the current tariff structure, househods consuming ess than 20 cubic meters per month (which is to say about 99% of househods in E Ato) pay a tariff of 42

45 5. CONSUMPTION DIFFERENTIALS Tabe 5-4 Average consumption for different user categories Observations Mean Standard 95% confidence interva error Lower bound Upper bound By type of househod Without sewerage With sewerage but without bathroom With sewerage and bathroom By presence of water coset Without With By presence of hand basin Without With By presence of shower Without With By presence of kitchen sink Without With By presence of aundry sink Without With Tabe 5-5 Resuts of regression mode for water consumption Coefficient Standard T-statistic Significant Consumption impact error at 95% eve? (m3/month) Habitation Househod size Yes m 3 /person Aways inhabited a Yes +1.74m 3 Sometimes inhabited a Services Sewerage Condominia water Water-reated instaations Bathroom WC b Sia Turca b Latrine b Hand basin Shower Yes +1.00m 3 Kitchen sink Yes +1.47m 3 Laundry sink Yes +1.08m 3 Neighborhood Huayna Potosí c Via Ingenio c Caja Ferroviaria c San Juan de Río Seco c Oro Negro c Jichusirca c Constant Yes Regression parameters Observations 1114 Adjusted R squared 0.32 F-statistic (19, 1094) a Defined reative to the baseine category of an unconstructed ot. b Defined reative to the baseine category of no toiet faciity of any kind. c Defined reative to the baseine category of the neighborhood 3 de Mayo, a contro area with no project interventions. 43

46 WATER AND SANITATION PROGRAM US$0.22 that amost certainy fas short of the average cost of service provision. Whie the true efficient average cost of the company is currenty unknown, estimates produced at the time of privatization put the vaue at US$0.56. Using this vaue purey as an iustrative point of reference, AISA coud be osing up to US$0.34 per cubic meter of consumption in the first consumption band. Hence, under the current regime, the increases in consumption produced by the hygiene education program because they are not arge enough to take househods into a higher tariff band ony serve to increase the osses that AISA experiences from connecting consumers in E Ato. If on the other hand, the tariff structure refected the true average cost of service provision, the hygiene education activities coud be expected to increase the utiity s revenues and hence potentiay profits by around 10%. However, without knowing the true average cost of the service or the utiity s profit margin and discount rate, it is not possibe to put an absoute monetary vaue on this benefit. Again, purey as an iustration, if the true average cost of water were US$0.56 and the company s profit margin and discount rate were 15%, a 0.7 cubic meter per month increase in consumption woud transate into a profit increase in perpetuity vaued at around US$5. Economic perspective Irrespective of the tariff structure, the hygiene education component eads to a change in consumer preferences that increases the amount of water demanded at every price. As a resut, househods experience an increase in consumer surpus. The monetary vaue of this increase depends on the price easticity of demand, a parameter that is unknown for the piot area. In order to take the anaysis further, some reativey heroic assumptions wi therefore have to be made about the functiona form of the demand curve. By identifying two points on the demand curve, it is possibe to recover the parameters of a inear demand function. The point representing current consumption of 6.6 cubic meters per month at a price of US$0.22 per cubic meter is aready known. Komives (1999) reports some approximate estimates of price and quantity of water used by consumers without piped suppies, finding that househods reying on tankered suppies pay US$3.64 per cubic meter and consume 2 cubic meters per month. If one is wiing to accept that these two points ie on the same demand curve and that the demand function is inear, then it is possibe to proceed with the consumer surpus anaysis 18. The monthy increase in consumer surpus for the inear demand curve defined by the above two points turns out to be US$2.74 (or US$32.92 on an annua basis) 19. In order to cacuate the wefare gain, it is necessary to take the present vaue of this change in consumer surpus. Unfortunatey, the discount rate of househods in E Ato is not known. However, it is known that ony a minority of househods were wiing to borrow money from CLA at 26% per annum, whie a majority were wiing to pay AISA an interest rate of 13% in order to spread the cost of connection charges. It therefore seems pausibe to say that househod discount rates ie coser to 13% than 26%. At a discount rate of 13%, the present vaue of the increment in consumer surpus in perpetuity works out at US$253. It is important to note that this is consideraby ower than the investment of US$441 required to construct a bathroom. Indeed, it was found that the consumer surpus from bathroom construction ony starts to exceed the start-up costs at discount rates beow 7.5%. As noted in passing, this anaysis is subject to numerous imitations and hence the resuts shoud be treated with considerabe caution. Perhaps the main message to be drawn is that the changes in consumer surpus appear to be of a simiar order of magnitude to the investments required to construct the bathroom, and hence that the concusion is highy sensitive to assumptions made about the discount rate over pausibe ranges of its possibe vaue. This suggests that from the househod s perspective bathroom construction represents neither a huge gain nor a huge oss, but rather something of a margina decision. This finding is in fact consistent with the ambivaence encountered among oca residents during the course of the hygiene education activities (reca Tabe 5-1). 44

47

48 WATER AND SANITATION PROGRAM 6. Evauation This section brings together a of the preceding resuts and uses them to conduct an evauation of the EAPP. As we as considering both the reative costs and benefits of condominia and conventiona systems, the margina costs and benefits of the various additiona components such as community participation and hygiene education are aso given separate consideration. The resuts are presented from the socia perspective, as we as the private perspectives of the utiity and the consumer. Socia perspective The evauation begins by considering the virtues of the condominia system from a socia perspective, that is to say, taking a costs and benefits into account regardess of the party to which they accrue. This perspective is appropriate to an enightened poicy maker and is hepfu in providing a hoistic picture of the condominia approach. The main resuts are summarized in Figures 6-1 and 6-2 (see Tabe A3 of the Annex for underying figures). The pattern is very simiar for both services. Network savings costs of around 50% are achieved, but these fa to 40% once the costs of socia intermediation and the opportunity cost of community time are fuy accounted for. Overa, the absoute saving is of the order of US$100 per connection. It is aso interesting to note that the opportunity cost of community time is more than twice as arge as the cost of socia intermediation, athough both are sma in reation to network expansion costs. These findings suggest that the condominia approach is very worthwhie from a socia perspective. In addition to making this overa comparison between conventiona and condominia systems, it is important to compare the variants of the condominia system with each other in order to estabish the baance of margina costs and benefits. Two comparisons are particuary important. First, it is interesting to examine the margina costs and benefits of the community participation component (Tabe 6-1). As noted above, community participation is responsibe for 49% of the network expansion savings for the sewerage service (equivaent to US$66 per connection) and 23% of the network expansion savings for the water service (equivaent to US$27 per connection). In the case of the sewerage service, these savings are arge in reation to the costs of socia intermediation and community participation, eading to a benefit cost ratio in 46

49 6. EVALUATION excess of two. This suggests that community participation is extremey worthwhie in the case of the sewerage service. The resuts for the water service are ess favorabe, since the benefits are smaer and barey offset the additiona costs, with a resuting benefit-cost ratio cose to unity. Figure 6-1 Socia evauation of condominia sewerage service The second comparison is between the versions with and without the hygiene education and reated bathroom support activities (Tabe 6-2). The hygiene education activities cost an additiona US$13 per househod to undertake, and take-up US$2 of househod time. The construction of the bathroom itsef costs US$441. Thus the overa cost of the investment is US$456. These activities were found to increase the proportion of househods constructing bathrooms by 35% and to increase water consumption in househods with bathrooms by about two cubic meters per month. Hence, the overa effect of this is to raise consumption by 0.7 cubic meters per month on average in the neighborhoods affected. Figure 6-2 Socia evauation of condominia water service Even assuming cost-refective tariffs, the present vaue of the profit margin on this additiona consumption amounts to very itte for the utiity. The gains in consumer surpus are arger by about two orders of magnitude. However, whether they are arge enough to compensate for the substantia costs of investing in a bathroom depends criticay on assumptions made about the discount rate. The overa concusion is that hygiene promotion is not at a attractive to the utiity, but is potentiay attractive to the househod. Given the exceptionay ow eves of water consumption in E Ato, it seems ikey that bathroom promotion campaigns in other ocations woud potentiay ead to much arger increases in househod water consumption. An increase of 10 cubic meters per month, for exampe, woud more than compensate the bathroom investment even at the 13% discount rate. It woud aso start to make the investment ook attractive to the utiity. Utiity s perspective The preceding section considered the comparison between conventiona and condominia sys- Tabe 6-1 Margina benefit-cost anaysis of community participation Benefits Costs Benefit-Cost Ratio Water Sewerage Tabe 6-2 Margina benefit-cost anaysis of hygiene promotion Benefits Costs Benefit-Cost Ratio Discount Rate 7% 13% US$ 7% 13% Customer Utiity 5* 5* Tota *This vaue is obtained by appying the average cost tariff of US$0.56 (estimated in 1997) to the expected consumption increment of 0.7m 3 per month, annuaizing it, and taking the present vaue in perpetuity. 47

50 WATER AND SANITATION PROGRAM Figure 6-3 Figure 6-4 Utiity s perspective on condominia sewerage service (US$) Utiity s perspective on condominia water service (US$) tems from a socia perspective. However, in order for a system to be impemented, it needs to be attractive not ony from a socia perspective, but from the narrower individua perspectives of the participating parties. To this end, this section adopts the perspective of a privatey managed utiity, assuming that the utiity is primariy concerned about the impact of aternative systems on profitabiity. The foowing concusions emerge. First, in practice, it is very difficut to assess the profit impact of making new connections. Athough the connection charges aowed in the concession contract are substantiay ower than the average costs of connection estimated in the present study (Figures 6-3 y 6-4), this does not necessariy mean that the utiity is faiing to recover the costs of connection since these may have been incorporated into the genera tariff eve. Second, athough part of the cost of network expansions may have been incorporated into the average user tariff, the very arge size of the subsistence bock means that new customers are not commerciay attractive because they fa in the owest tariff band or US$0.22 per meter, compared with the ast pubished estimate of the average cost of service which was US$0.56 per cubic meter (Uzin, 1996), the higher consumption eves of condominia customers cost the utiity US$36 per year but bring in ony US$14 per year of revenue since they fa in the owest tariff band of US$0.22 per cubic meter. Hence, serving the newy connected customers is a oss-making proposition. Third, under the current tariff structure, there is no incentive whatsoever for the utiity to undertake the hygiene education component. This is so for two reasons. On the one hand, hygiene education increases the costs of connection by US$13 without any concomitant increase in the connection charge. On the other hand, athough hygiene education eads to higher water consumption the increases are not arge enough to take househods into a higher tariff band; and higher water consumption in the first tariff band ony eads to higher osses. These concusions woud change competey if the tariff structure were brought in ine with the underying structure of costs. If this were the case, the utiity woud be indifferent to the cost savings of the condominia approach since the profitabiity of both types of connections woud by design be identica. Furthermore, it woud aso be profitabe to provide service to ow voume consumers. However, as noted above, due to the reativey sma increases in consumption resuting 48

51 4. COST DIFFERENTIALS from the hygiene education activities, these woud remain unattractive to the utiity uness the associated costs coud be recovered from the consumer in the connection charge. Figure 6-5 Customer s perspective on condominia sewerage service (US$) Consumer s perspective This section adopts the perspective of a consumer facing a choice between conventiona and condominia systems, and between aternative variants of the condominia system (see Figures 6-5 y 6-6 and underying figures in Tabe A5 of the Annex). Separate resuts are reported for the current structure of charges and a hypothetica aternative where charges refect the underying structure of costs. The main concusions are as foows. First, the current tariff structure offers the consumer no cost advantage from the condominia water service and substantiay attenuates the cost advantage avaiabe from the condominia sewerage service. The tota savings currenty avaiabe to househods, at US$58 are nonetheess significant representing 50% of the average monthy househod income in E Ato. Figure 6-6 Customer s perspective on condominia water service (US$) Second, if charges for condominia systems were made fuy cost-refective, househods woud stand to make significant gains. The overa saving on the water system woud be cose to 40% for both water and sewerage services. The tota vaue of the savings from opting for condominia water and sewerage connections woud come to US$194, which is equivaent to 160% of the average monthy househod income in E Ato. EL ALTO - BOLIVIA / PILOT PROJECT 49

52 WATER AND SANITATION PROGRAM 7. Concusions This paper undertakes the economic and financia vauation of the EAPP, a piot project aimed at transferring the condominia water and sewerage system from Brazi to Boivia and testing its appicabiity in the context of private sector participation in service provision. Since its inception in 1998, the EAPP has provided condominia water connections to 1,977 househods in eight neighborhoods of E Ato, and condominia sewerage connections to 4,050 househods in nine neighborhoods of E Ato. According to recent statistics, about 60% of these househods ive beow the poverty ine with an average daiy income per capita of US$0.80. degree. Exampes incude the exceptionay ow eves of consumption by househods, and the difficuties experienced in inducing them to switch to modern hygiene practices, for a variety of cutura reasons. Consequenty, the resuts of the evauation shoud be regarded as specific to the E Ato context, even though in quaitative terms they are indicative of what can generay be achieved through the condominia approach. The piot project experimented with a number of different components incuding: - innovative engineering design of networks; - community participation in network construction and maintenance; - hygiene education to support the instaation of househod faciities; - micro-credit ines to finance the construction of bathrooms. To what extent are the resuts of the E Ato piot reevant beyond Boivia? The pecuiar cutura, geographica and socia circumstances of E Ato make it if anything an acid test for the condominia approach. In particuar, a number of factors which have served to imit the benefits of the condominia system in Boivia woud not necessariy be present in other contexts to the same Many of these components are mutuay reinforcing and formed part of an integrated concept in the E Ato project. Nonetheess, both in principe as in practice, it is possibe to appy these components independenty of each other. Hence, for anaytica purposes it proves convenient to provide an independent evauation of each. 50

53 7. CONCLUSIONS Engineering design The purpose of the innovative engineering design is to reduce the ength, diameter and depth of the network required by routing the distribution pipes across pavements and/or backyards. Anaysis of the EAPP experience suggests that savings in the ength and diameter of pipes are of the order of 10%-20%, whie savings in the voume of soi excavation as a resut of shaower trenches are of the order of 45%-75%. These physica savings transate into overa financia savings of the order of 24% for the sewerage service and 40% for the water service when the condominia engineering design is impemented using conventiona contractors. This is consistent with recent experience from Brasiia which suggests savings of 20% for condominia sewerage systems without community participation (Neder, 2001). Community participation Community participation brings a number of advantages, among them is a further reduction in connection costs as a resut of training oca residents to construct and maintain their own condominia branches. Community participation reduced the network costs by a further 26% for the sewerage service and 10% for the water service. Thus, the overa savings achievabe by impementing the condominia engineering design with community abor come to around 50% for each of the two services. However, community participation aso introduces costs of socia intermediation for the water company of around US$8 per connection, and requires each participating househod to give-up about a week of its time vaued at around US$20. When these costs are taken into account, the overa cost advantage of the condominia design with community participation fas sighty from 50% to 40% for both services. Considering the baance between margina costs and benefits, community participation has a benefit cost ratio of 2.2 for the sewerage service suggesting that it is very worthwhie. However, in the case of the water service the benefit cost ratio is ony 1.0, since in this case the cost savings come primariy from the engineering design. Figure 7-1 Cost comparison (US$) There is aso some evidence suggesting that community participation increases the proportion of househods that connect to the sewerage network once it is buit from 66% to 75%. The effect of this is to increase the overa cost saving of the approach from around 40% to 45%. Hygiene education The purpose of the hygiene education component was to provide mora and technica support for househods to adopt modern hygiene practices, in particuar by heping them to construct their own bathrooms and associated faciities. Without such investments within the home, a sewerage connection brings itte or no benefit to househods, and has been shown to have virtuay no impact on water consumption. In neighborhoods where hygiene education activities had been performed the probabiity that any given househod had constructed a bathroom increased from 38% to 73%, and the range of water-using instaations in the househod (showers, kitchen sinks, etc.) increased markedy. Househods with bathrooms were found to consume, on average two cubic meters more per month than househods without bathrooms, which is to say an increase of 30% on average consumption in househods without bathrooms of 5.4 cubic meters per month. Househods receiving hygiene education were aso found to significanty reduce insanitary practices (Canei, 2001). The percentage throwing out used water into the streets fe from 77% before the EL ALTO - BOLIVIA / PILOT PROJECT 51

54 WATER AND SANITATION PROGRAM project to 58% thereafter. Whie the percentage recycing water within the home fe from 36% to 25%. The cost of hygiene education to the utiity works out at around US$13 per househod, whie the cost of bathroom construction to the househod was estimated at US$443. Given the reativey sma absoute vaue of the consumption increase, and the ow eve of tariffs for ow voume consumers, hygiene education is not at present a commerciay attractive proposition for a private utiity in E Ato. The estimated consumer surpus increase from the additiona consumption ranges from US$253-US$470 depending on the discount rate used. This suggests that bathroom construction is ony marginay attractive to househods. However, it is important to stress that these modest benefits for hygiene education are argey an artifact of the exceptionay ow eves of water consumption in E Ato. With the much arger consumption increases that coud easiy be anticipated in other ocations, the proposition woud start to ook much more attractive both from the utiity s and the customer s perspectives. Micro-credit ine The purpose of the micro-credit ine was to hep househods finance the US$400 worth of materias required to construct a fuy-equipped bathroom. Overa, 25% of househods appied for credit, and 19% had their appications approved. The avaiabe evidence suggests that those househods appying for credit tended to have above average incomes. Whie the provision of credit undoubtedy eased the payment of the costs for beneficiary househods, statisticay there is no evidence to suggest that bathroom construction rates were significanty higher in neighborhoods were micro-credit had been offered. The microcredit ine was dropped at an eary stage of the project. Overa, these resuts suggest the foowing concusion. The engineering design component aone produced significant savings of 24% for sewerage and 40% for water. The community participation component further increased the savings avaiabe for the sewerage service to 40%, but did not have any net effect on the cost reductions avaiabe for the water service. The hygiene education component had a very substantia effect on househod behaviour, but the size of the benefits was more modest than might be expected owing to very abstemious water consumption practices pecuiar to E Ato. Evidence on the micro-credit faciity suggests that it probaby was not a that effective in reaching the poorest househods. The resuts reported above are given from the perspective of an enightened poicy maker who takes a financia and economic costs fuy into account. However, the considerabe divergence between tariff structure and underying cost structures in the AISA concession contract appears to distort the way in which the utiity and its consumers perceive the advantages of the condominia system. From the utiity s perspective, it is difficut to make categorica statements about the profitabiity of making condominia as opposed to conventiona connections. Athough current connection charges mandated by the concession contract are substantiay ower than the estimated average cost of making a connection. It is not known to what extent part of the costs of network expansion have in fact been buit into the use of service tariff. However, one point that is cear is that there does not appear to be any incentive for the utiity to engage in hygiene education since the additiona costs are not recouped through the connection charge, and the resuting increases in water consumption though significant are not arge enough to take consumers out of the oss-making ow voume tariff band. If charges were fuy cost-refective the utiity woud be indifferent between conventiona and condominia systems, but woud sti ony have an incentive to undertake hygiene education as ong as the consumption increase was arge enough to justify the cost of the initia investment. From the consumer s perspective, househods receiving water and sewerage services by the condominia method saved US$19 and US$80 52

55 7. CONCLUSIONS respectivey in terms of connection charges, an aggregate vaue equivaent to 80% of monthy househod income. However, when the opportunity cost of the househod s time is fuy accounted for, the saving disappears for the water service and for the sewerage service is reduced to US$58 (or about 50% of monthy househod income). If connection charges were more cosey aigned with underying cost, househods woud spend more on obtaining a condominia connection than they do at present, but the savings reative to the conventiona system woud aso be arger. They woud increase to US$109 for water and US$126 for sewerage (together equivaent to neary 200% of monthy househod income). As before, when the opportunity costs of time are taken into account, the savings are somewhat reduced to US$90 for water and US$104 for sewerage (in tota equivaent to 160% of househod income).

56 Notes 1 Loca term referring to an arid high atitude pateau. 2 In Spanish, this approach is described as con gestión compartida. 3 In Spanish, this approach is described as sin gestión compartida. 4 This is based on data for average earnings of workers in E Ato from the 1999 Encuesta de Hogares, which are just under $100 per month. The average number of hours worked per week is 50, which is taken to represent 6 days of work per week, or 24 days over the month. Hence, the average daiy wage is just over $ This probem affected the projects at Via Ingenio (UV-4), E Rosa Loreta and Kupini II. 6 The average reates to the foowing neighborhoods: Ata Ciudadea (Parts I and II), Huacataqui, and Bartoina Sisa, This average excudes one exceptionay difficut neighborhood, Peña Azu, where the cost per connection proved to be as high as US$811. Incuding this neighborhood in the cacuation, raises the average cost per connection to US$ This breakdown is made by considering the roes of different team members. The training of oca professionas in the design and impementation of condominia systems was argey undertaken by Adea Martínez, Mery Quitón, Fernando Inchauste and Afonso Nueva. Whereas, the adaptation of the approach together with the training of the trainers was undertaken primariy by Luiz Lobo (Team Leader) 8 However, non-owner occupying househods aso took out smaer oans. The average oan for this group was US$366, putting the principa asset ratio for this category at 64.2%. 9 This comment is based on an interview with Caudio Parra, Regiona Manager for CLA in E Ato. 10 This comment is based on an interview with Pedro Arrioa, Genera Manager of CLA. 11 To fuy remove the effects of seasonaity a fu year s average before and after woud have been idea. However, there was not sufficient data after the event to permit this. By ooking at the same six month period, at east one can be sure that the differentia is due to the presence of a bathroom rather than to varying consumption patterns at different times of year. Nonetheess, the size of the differentia wi not necessariy be constant across the year. 12 This can be inferred from the fact that the confidence interva for the difference between consumption before and after the arriva of sewerage spans the vaue zero. 13 This can be inferred from the fact that the confidence interva for the difference between consumption before and after the arriva of sewerage does not span the vaue zero. 14 The confidence intervas for the two consumption eves amost overap. The upper bound for househods without sewerage is 4.94 cubic meters per month, whie the ower bound for househods with sewerage but without bathrooms is This difference is just significant at the 95% confidence eve, but woud not be at the 99% confidence eve. 15 This can be inferred from the distance that exists between the upper bound of the first confidence interva (5.85) and the ower bound of the second confidence interva (6.86). 16 This figure can be arrived at in a number of ways. First, the rate of bathroom construction is 35% higher when hygiene education is appied, and the presence of a bathroom on average adds 1.9 (cross-sectiona) to 2.0 (time-series Huayna Potosí) cubic meters per month to consumption. Thus, the overa effect is (0.35*1.95) = 0.68 cubic meters per month. Aternativey, the instaations that add significanty to water consumption are the shower, the aundry sink and the kitchen sink. The coverage of these instaations is respectivey 24%, 33% and 12% higher in neighborhoods receiving hygiene education, and the consumption differentia is on average 1 cubic meter, 1 cubic meter and 1.5 cubic meters. Thus, the overa effect is (1*0.24+1* *0.12) = 0.75 cubic meters per month. This resut is consistent with the preceding one. 17 That is to say that in about 35% of househods consumption rises by about 30%, so that overa consumption rises by about 10%. 18 Evidenty, both of these assumptions are highy questionabe. On the one hand, a tankered suppy is so different in nature from an inhouse connection that the two points are unikey to be on the same demand function. On the other hand, there is no reason why the demand function shoud be inear; athough, neither is there any a priori reason to expect it to be anything ese. 19 The inear demand curve resuting from these two points has an easticity of ony 0.04 at the current price of water. Since this vaue is impausiby ow, some sensitivity anaysis was performed by owering the price of tankered water unti the price easticity rose to a more typica range ( 0.3 to 0.7). Fortunatey, the increase in consumer surpus is very robust to these variations, remaining in the interva $2-$3 per month. 54

57

58 WATER AND SANITATION PROGRAM Bibiography Bakaian, A., Wright, A., Otis, R. and de Azevedo Netto, J. (1994) Simpified Sewerage: Design Guideines, UNDP-Word Bank Water and Sanitation Program. Cannei, N. (2001) Proyecto Pioto Condominia E Ato-La Paz: Impacto de Proyecto, Session at Workshop: Presentación de os Resutados de Impantación de Sistema Condominia La Paz y E Ato, Ministerio de Vivienda y Servicios Básicos, Agencia Sueca de Cooperación Internaciona para e Desarroo, Programa de Agua y Saneamiento, Aguas de Iimani, La Paz. Komives, K. (1999) Designing Pro-Poor Water and Sewer Concessions: Eary Lessons from the Aguas de Iimani Concession in Boivia, Mimeo, Private Participation in Infrastructure Group, Private Sector Deveopment Division, Word Bank, Washington DC. Luduvice, M., Neder, K.D., Teixera Pinto, M. (2001) Produtividade e Eficiência na Impantação e Operação de Sistemas de Esgotamento Sanitário A Experiência da CAESB, Mimeo, Brasiia, Brazi. Neder, K. (2001) Sistema Condominia: Aspectos Economicos y Financieros de a Experiencia en e Brasi, Session at Workshop: Presentación de os Resutados de Impantación de Sistema Condominia La Paz y E Ato, Ministerio de Vivienda y Servicios Básicos, Agencia Sueca para e Desarroo Internaciona, Programa de Agua y Saneamiento, Aguas de Iimani, La Paz. Quisbert, T.R. (2000) Concepción e Impantación de Sistemas Condominiaes de Acantariado Sanitario y Agua Potabe en a Ciudad E Ato, Tesis de Licenciatura, Carrera de Ingeniería Civi, Facutad de Ingeniería, Universidad Mayor de San Andrés, La Paz, Boivia. Mara, D. (2001) Introduction to Low-Cost Sewerage, Sanitation Connection Website ( Navajas, S., Schreiner, M., Meyer, R.L., Gonzáez-Vega, C. and Rodríguez-Meta, J. (2000) Microcredit and the poorest of the poor: Theory and evidence from Boivia, Word Deveopment, Vo. 28, No. 2, pp Programa de Agua y Saneamiento (2001) Procedimientos de Operación y Mantenimiento de Sistema Condominia, Mimeo, Programa de Agua y Saneamiento, La Paz, Boivia. Uzin, L.G. (1996) Estudio Tarifario SAMAPA, Mimeo, Consutant s Report to SAMAPA, La Paz. Watson, G. (1995) Good Sewers Cheap? Agency-Customer Interactions in Low-Cost Urban Sanitation in Brazi, Water and Sanitation Currents, UNDP-Word Bank Water and Sanitation Program. 56

59

60 WATER AND SANITATION PROGRAM Cost Annex Tabe A1 Financia costs per sewerage connection Cost (US$) Cost difference Condominia Conventiona* Absoute Reative ower midde upper ower midde upper (US$) (%) Phase One Huayna Potosí Via Ingenio (D-II) Caja Ferroviaria San Juan de Río Seco Oro Negro Phase Two E Ingenio Germán Busch* Overa average *It is important to note that the numbers in these rows and coumns are budgetary estimates as opposed to outturn costs. Tabe A2 Financia costs per water connection Cost (US$) Cost Difference Condominia Conventiona* Absoute Reative ower midde upper ower midde upper (US$) (%) Phase One Via Ingenio (D-II) Caja Ferroviaria San Juan de Río Seco Jichusirca Phase Two Germán Busch* Averages *It is important to note that the numbers in these rows and coumns are budgetary estimates as opposed to outturn costs. 58

61 Tabe A3 Evauation resuts from a socia perspective Absoute Vaues Water Sewerage Conventiona Condominia Conventiona Condominia Costs (US$) - Network expansion Pus socia intermediation Pus opportunity cost of time Absoute Differences Cost savings (US$) - Network expansion Pus socia intermediation Pus opportunity cost of time Percentage Differences Cost savings (%) - Network expansion Pus socia intermediation Pus opportunity cost of time Tabe A4 Evauation resuts from the utiity s perspective Water Sewerage Conventiona Condominia Conventiona Condominia Connection Costs Revenues 119* 100* Profit impact Consumption (annua) Costs* Revenues Profit impact *This is the US$120 financia cost of the connection minus the US$36 cost of the meter. These are based on the average cost tariff of US$0.56 (estimated in 1997). EL ALTO - BOLIVIA / PILOT PROJECT 59

62 WATER AND SANITATION PROGRAM Tabe A5 Resuts of evauation from the consumer s perspective Absoute Vaues Water Sewerage Conventiona Condominia Conventiona Condominia Actua charges - Monetary cost 119* 100* Fu cost Cost-refective charges - Monetary cost Fu cost Absoute Differences Actua charges - Monetary cost Fu cost 0-58 Cost-refective charges - Monetary cost Fu cost Percentage Differences Actua charges - Monetary cost Fu cost 0-32 Cost-refective charges - Monetary cost Fu cost *The US$36 cost of the meter has been subtracted from the current connection charge in order to make this comparison consistent. 60

63

64