WATER AND SANITATION PROGRAM: GUIDANCE NOTE 67393 Sanitation Finance in Rural Cambodia Andy Robinson February 2012 The Water and Sanitation Program is a multi-donor partnership administered by the World Bank to support poor people in obtaining affordable, safe, and sustainable access to water and sanitation services. Acknowledgments The Program would like to thank the Asian Development Bank for its valuable technical and financial contribution. This guidance note reported here would have been impossible to put together without the active help and cooperation of the people managing the projects on which these studies are based. The author gratefully acknowledges Dr. Mao Saray and Dr. Chea Samnang (Ministry of Rural Development), Wan Maung (Tonle Sap Rural Water Supply and Sanitation Project), Cordell Jacks (International Development Enterprises), Lyn McLennan (WaterSHED, previously Lien Aid), Aun Hengly (WaterSHED, previously World Toilet Organization), Danielle Pedi (IDE, previously World Toilet Organization), Syvibola Oun (PLAN International), Geoff Revell (WaterSHED) and staff of the Tonle Sap Rural Water Supply and Sanitation Project Provincial Management Units in Siem Reap and Kampong Thom. Thanks also to Karin Schelzig-Bloom (Asian Development Bank) and Jan Willem Rosenboom (Bill and Melinda Gates Foundation, previously Water and Sanitation Program) for their help with the study and comments on successive drafts of this report. Water and Sanitation Program (WSP) reports are published to communicate the results of WSP’s work to the development community. Some sources cited may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein are entirely those of the author and should not be attributed to the World Bank or its affiliated organizations, or to members of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank Group concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to wsp@worldbank.org. WSP encourages the dissemination of its work and will normally grant permission promptly. For more information, please visit www.wsp.org. © 2012 Water and Sanitation Program Sanitation Finance in Rural Cambodia Andy Robinson February 2012 Sanitation Finance in Rural Cambodia Executive Summary This document presents the findings of a study on sanita- The document ends with recommendations for improved tion finance in Cambodia conducted for the Water and sanitation finance, including practical suggestions for sani- Sanitation Program (WSP) with support from the Asian tation programs in Cambodia. These recommendations Development Bank (ADB). The overall objective of the as- bear particular relevance for the ADB’s Second Rural Water signment was to consider sustainable sanitation financing Supply and Sanitation Sector Project, which commenced options with a focus on promoting access for the poorest. in 2010. This guidance note contains an introduction on sanitation PUBLIC FINANCE OF SANITATION financing and subsidies, stating the cases for subsidies as Sanitation improvement provides a particular development well as some of their practical pitfalls. The study used data challenge due to the difficulty of generating private demand (as of late 2009) from two case studies of rural sanitation for sanitation facilities. Awareness of the private and exter- finance in Cambodia to illustrate the practical issues, sup- nal costs of inadequate sanitation is generally low in devel- plemented by preliminary data from two sanitation mar- oping countries. Despite widespread diarrheal disease and keting projects. The study also examined the potential use high child mortality rates, health costs are rarely ascribed to and effectiveness of (hardware) subsidies, conditional cash unhygienic sanitation practices, toilets are often perceived transfers (CCTs), and other financing approaches relevant to be unaffordable, and demand for improved sanitation for sanitation improvement. remains low. www.wsp.org iii Sanitation Finance in Rural Cambodia | Executive Summary Everyone without an improved sanitation facility is “sanita- The exact relationship between health outcomes and sani- tion poor� and, therefore, will benefit from public support tation status remains uncertain, but children from poor to improve sanitation. Furthermore, each new improved households have significantly higher mortality and mal- sanitation facility will reduce the number of pathogens in nutrition rates than those from non-poor households; and the environment, thus provide societal as well as private the risk of diarrheal disease in Cambodia is three to four benefits. In a context of low demand for sanitation, this times higher among severely underweight children. There- framework suggests that there is little need for targeting fore, children from poor households, particularly those that (among those not using improved sanitation facilities) as are malnourished, are likely to contribute more pathogens any new toilet will be beneficial. to the environment through unsafe excreta disposal than children from non-poor households. As a result, sanitation The best welfare-enhancing approaches would target the strategies that fail to deliver improved sanitation to poor poorest first, due to the higher marginal value of each households are likely to have less optimal outcomes, with monetary unit among the poorest. But, in practice, sanita- fewer health and economic benefits, than those that succeed tion programs often target the “low-hanging fruit� – those in reaching the poor. without improved sanitation facilities that are more willing to invest, more responsive to promotional programs, and TYPE OF BENEFIT easier to reach. The intention of this targeting is that, in The social welfare literature suggests that cash transfer is addition to the benefits from the additional sanitation fa- the best instrument for addressing concerns about poverty cilities, the supply of sanitation goods and services to these and inequality because it respects the principle of consumer responsive households will build a larger sanitation market, sovereignty, allowing beneficiary households to allocate the thus developing the economies of scale and common good additional income they receive to the good or service repre- practice that will be needed to change sanitation behavior senting the highest priority for the household. Cash trans- and spending priorities among the poorest households. fers also avoid potential distortions in economic decisions caused by changes in the relative prices of goods. In Cambodia, where 77 percent of the rural population lack improved sanitation and 35 percent of rural households are Despite the theoretical advantages of cash transfers, the below the official poverty line, more than half of those with- prevalent instruments of social policy and sanitation fi- out improved sanitation facilities are non-poor households. nance in most developing countries are in-kind transfers. As a result, sanitation programs that target “low-hanging The practical concerns associated with cash transfers in fruit� tend to benefit largely non-poor households because developing countries, such as administrative constraints many sanitation programs require a household contribution linked to limited financial infrastructure, inefficient target- that is unaffordable or unattractive to poor households. ing mechanisms, and the risk of corruption, result in most programs utilizing in-kind transfers. In addition, there is no guarantee that cash transfers will be used to improve human capital unless specific conditions are attached to the use of the transfers, which further complicates the administrative requirements. Recent growth in the use and effectiveness of conditional cash transfers demonstrates that the administrative require- ments are no longer insurmountable in developing coun- tries. Technological advances mean that the options for financial payment, means testing and compliance monitor- ing systems have improved dramatically; while there is in- creasing evidence of the drawbacks of conventional in-kind subsidies. iv www.wsp.org Sanitation Finance in Rural Cambodia | Executive Summary MINIMUM BENEFIT LEVEL SERVICE DELIVERY The administrative cost associated with any benefit transfer Recent research by the IDE Cambodia sanitation market- system – for operation of the targeting, delivery and moni- ing project (supported by WSP and USAID) found that toring systems – limits the minimum effective size of the local producers in two provinces were willing to provide the benefit. Significant economies of scale are available in the basic below-ground components of a pour-flush latrine for administration of national benefit programs, and additional a total cost of only US$25. Yet a similar package of latrine economies are available from the use of existing targeting materials provided by private contractors through the ADB and financial systems. Nevertheless, the benefit level needs Tonle Sap Rural Water Supply and Sanitation Program to be set high enough to justify the administrative expen- (TS-RWSSP) currently costs more than US$88 per latrine. ditures. Wherever possible, sanitation programs should aim to sup- It is also important that the benefit level is set high enough port the provision of services through local markets rather to have an impact on the beneficiaries’ behavior and invest- than through parallel project delivery systems. The inten- ment decisions. A recent review of conditional cash trans- tion is to reduce the costs of supply, to provide the user with fer (CCT) programs found that monthly benefit payments a choice of service providers, to make the service provider ranged between 1 percent and 29 percent of pre-transfer accountable to the user of the services (rather than to the household expenditures. In Cambodia, this implies a CCT financier or project manager), and therefore to encourage benefit level of US$1 to US$20 per month, or as much as competition and service improvements among suppliers US$240 per household per year, which is higher than the and service providers. transfer provided by most sanitation programs. CONDITIONAL CASH TRANSFERS FOR However, improvements in sanitation behavior take time, SANITATION IMPROVEMENT thus there is an argument for the provision of benefits Although a relatively recent phenomenon, the CCT litera- over a longer time period in order to achieve sustained ture contains a number of evaluations that highlight the improvements in behavior. The first step towards im- success of CCT programs in improving the uptake of health proved sanitation is to stop open defecation through a and education services such as preventive health checkups, change in individual sanitation behavior. Sanitation finance vaccinations, and school enrolment. However, the evidence mechanisms should recognize that sanitation improvement that these gains result in improvements in health and edu- involves a number of different steps over time, includ- cation outcomes is mixed, with recent evaluations reporting ing initial behavior change, construction of a sanitation little or no improvement in malnutrition rates and learning facility, adoption of improved sanitation behaviors, im- outcomes. provement of the sanitation facility, and safe disposal of wastes. Recent research has confirmed that inadequate sanitation plays a significant role in the nutritional status of children. Therefore, while sanitation policies and strategies should The intuition that diarrheal disease caused by inadequate promote universal use of improved sanitation facilities, san- water supply and sanitation affects nutritional uptake, and itation finance should be designed to promote and support that malnutrition in turn increases the relative risk of di- the adoption of a range of sanitation behaviors, and a range arrheal disease, is supported by a recent collective expert of sanitation technologies that will change as individuals opinion that about 50 percent of the consequences of mal- and communities become more familiar with improved nutrition are caused by inadequate water and sanitation sanitation behavior and more willing to invest in improved services and poor hygienic practices. Repeated infections, sanitation services. Importantly, the setting and financing especially diarrhea and helminthes, caused by poor environ- of any minimum level of sanitation service should not pre- mental health lead to underweight (low weight for age) and clude or prevent the construction of low-cost hygienic fa- stunted (low height for age) children, and these conditions, cilities, or constrain the development of innovative, local in turn, make these individuals more predisposed to preda- latrine designs. tory infections and chronic diseases later in life. www.wsp.org v Sanitation Finance in Rural Cambodia | Executive Summary Most CCT nutrition programs target mothers with young 25 percent of the total population of Cambodia. The proj- children, with regular payments made based on records of ect villages will include about one in four of all villages health and nutrition service use (including child growth within the five provinces, thus the project aims to make monitoring, vaccinations and attendance at nutrition semi- a substantial impact on quality of life and environmental nars). Despite increased awareness of the links between health within these provinces. malnutrition and diarrhea, few nutrition programs include any components that promote improved sanitation and hy- Expenditure data collected by the WSP ESI study suggest giene. Initial discussions suggest that the current failure to that the typical user contribution per ADB pour-flush la- link sanitation improvement and nutrition provides a sig- trine is as follows: nificant opportunity for the improvement of CCT nutrition • Cash contribution = US$90 programs, through the potential for additional conditions • In-kind materials = US$46 that encourage the use of improved sanitation facilities and • Labor = US$1 the achievement of collective sanitation outcomes. Total household = US$137 per latrine CASE STUDIES In 2007, data collected by the WSP-IDE sanitation de- Four case studies were completed as part of the research for mand assessment study suggested that 73 percent of rural this study. The first two were detailed case studies of pro- households (that do not own a latrine) were willing to pay grams that had already been running for three years (as of US$10 for a pit latrine. The same data suggest that just over 2009 where the study was undertaken), using preliminary 40 percent would be willing to pay US$30 for a pit latrine household survey data collected by WSP’s Economics of (the cost of the “sanitation core� currently being promoted Sanitation Initiative (ESI) Phase 2 study: by the IDE sanitation marketing program), and that only 15 percent would be willing to pay the US$80 contribution • ADB Tonle Sap Rural Water Supply and Sanitation demanded by the ADB TS-RWSSP. Project (ADB TS-RWSSP); • Plan Cambodia CLTS program. The project expenditure data show that three-quarters of the sanitation investment to date has been on pour-flush la- The other two case studies were partial assessments of sani- trines, all of which are likely to have been built by non-poor tation marketing interventions1: households that could afford the US$80 cash contribution; while at most half of those that built dry latrines are likely • IDE sanitation marketing project; to be poor households (as the level of investment made in • WTO and LienAid sanitation marketing project. building the latrine superstructures indicate that the major- ity were non-poor households). In total, only about 10 per- The information on the case studies was drawn from ESI cent of the sanitation investment appears to have reached household surveys (where available), interviews with pro- those below the poverty line; the remaining 90 percent has gram managers and technical advisers, data provided by the benefited non-poor rural households that took advantage implementing organizations, and analysis conducted by the of the generous subsidies offered by the ADB TS-RWSSP. author. PLAN CAMBODIA COMMUNITY-LED TOTAL ADB TONLE SAP RURAL WATER SUPPLY AND SANITATION PROGRAM SANITATION PROJECT The Plan Cambodia Community Led Total Sanitation The TS-RWSSP is the largest rural sanitation program in (CLTS) program was initiated as a pilot project in two vil- Cambodia, with an estimated US$5.1 million spent on the lages in 2006. Since then, the program has expanded to sanitation component over the last three years. The project cover another 45 villages with a total expenditure of just un- area includes 3.42 million people, who comprise roughly der US$0.5 million. The Plan interventions aim to achieve 1 At the time of the study in late 2009, these sanitation marketing programs were just recently launched. vi www.wsp.org Sanitation Finance in Rural Cambodia | Executive Summary open defecation free communities, where 100 percent of The main implementation phase was launched in late 2009, the population use sanitation facilities. As in most CLTS immediately prior to the fieldwork for this study, thus the programs, no hardware subsidy or in-kind benefits are pro- review herein is based largely on proposed implementation vided to households that build private sanitation facilities. plans rather than an evaluation of project outcomes. The ESI household survey data suggest that, while signifi- Willingness to pay data from an IDE study completed in cant numbers have built latrines without any external sub- 2007 suggest that less than 43 percent of households with- sidy, latrine coverage and usage levels are low. Latrine usage out a latrine will be willing to pay this much – US$30 or (among those that own latrines) is estimated at 64 percent, more – for a new latrine. Given that roughly half of those based on the number of non-operational latrines reported without latrines are below the poverty line, these data imply in the ESI survey villages. Therefore, 41 percent of the proj- that few poor households will be willing or able to afford ect population are using their latrines, with another 23 per- the IDE latrine core package. cent having abandoned their latrines, and the remaining 35 percent with no latrines. However, IDE reports that its initial village marketing cam- paigns have generated sales right across communities, with The 2009 MRD CLTS evaluation in Cambodia found that even poor households committing to invest US$25 in the a similar proportion of simple CLTS latrines had been aban- latrine core package (with another US$5 to be spent on the doned during the rainy season due to flooding or collapse. enclosure). Willingness to pay estimates are notoriously un- However, the MRD evaluation also reported that most of reliable, thus it is possible that effective marketing of a more the people who had abandoned their latrines were practic- desirable sanitation product has shifted the willingness to ing “dig and bury� techniques during the rainy season and pay curve upwards, capturing a far higher proportion of planned to return to latrine use once the rains were finished those without latrines at the US$25 price point. and they were able to clean and repair their latrines. The program cost data suggest that the following average More than two-thirds of the Plan latrine-owning house- costs will be linked to the program sanitation development holds reported that they had not used any cash in the con- activities: struction of the latrine, with the total latrine cost valued Software (hygiene promotion, at only US$11, including all labor and in-kind materials. training) = US$20 per latrine The remaining households with self-built toilets reported Program costs (management, technical assistance) = US$56 per latrine an average cash expenditure of US$6 on top of similar in- Average project sanitation cost = US$ 76 per latrine kind material and labor contributions, making a total cost of US$17. Therefore, the average amount spent on a latrine Household contribution (cash and in-kind) = US$60 per latrine in the Plan households surveyed was less than 10 percent of Average total latrine cost = US$136 per latrine the amount spent on the ADB latrines. IDE SANITATION MARKETING PROJECT WTO-LIENAID SANITATION MARKETING The project is supported by both WSP and USAID, with a PROJECT combined budget of US$760,000 over a 21-month period. The World Toilet Organization (WTO) and LienAid sani- The project target is to sell 10,000 household latrines in tation marketing project builds on the research and devel- two provinces using a market-based approach with no sani- opment work done by the IDE project, supplemented by its tation hardware subsidy. The R&D work enabled IDE to own research on reasons for investment in sanitation facili- identify strong demand for a package of latrine components ties. One of the key differences between the two sanitation that enables rural households to build their own pour-flush marketing projects is that the WTO-LienAid project has latrine, including a basic enclosure, for as little as US$30. identified a durable superstructure as critical to the success www.wsp.org vii Sanitation Finance in Rural Cambodia | Executive Summary of rural sanitation interventions, thus the project would households. It is important that an appropriate amount of also market flat-pack latrine enclosures. public finance is directed towards developing and market- ing products and services that are specifically targeted at the As of the time of the study in 2009, the WTO-LienAid poorest households and those that cannot afford the US$30 program is not yet in its full implementation phase, but it sanitation core package. planned to sell 4,000 latrine packages through trained pro- ducers over the following 12 months. Given a total budget Finally, few of the programs examined have been successful of US$338,000, achievement of this target will result in a in achieving collective sanitation outcomes, such as open program cost per latrine of about US$84. defecation free communities, which should be the ultimate aim of all sanitation programs (in order to achieve the opti- The cost data suggest that the following average costs will be mal benefits). The population segment that practices open linked to the project sanitation activities: defecation in the program communities is largely made up Software (marketing, research, of poor households, and generally includes those with the strategy) = US$28 per latrine highest disease burdens – those that are most likely to trans- Program costs (management, mit diseases to others through unsafe excreta disposal. As a technical assistance) = US$56 per latrine result, the benefits achieved by these sanitation programs Average project sanitation cost = US$ 84 per latrine may be limited. Household contribution (cash and in-kind) = US$80 per latrine LESSONS AND RECOMMENDATIONS Average total latrine cost = US$164 per latrine 1. Leave no one out: complementary programs are needed to reach all groups COMPARATIVE ANALYSIS OF CASE STUDIES 2. Check who benefits: monitor targeting effectiveness The comparative analysis confirms that public finance for 3. Aim for efficiency: recognize the advantages of mar- sanitation in Cambodia is not reaching those below the ket-led service delivery poverty line. Ninety percent of the public finance for the 4. Use vouchers to encourage sustainable service provi- large ADB program goes to non-poor households, and the sion two sanitation marketing projects will require households 5. Design finance for long-term improvements in sani- to contribute at least US$30 in order to obtain a latrine, tation practice whereas the willingness to pay data imply that US$10 is the 6. Use large-scale means testing systems wherever pos- maximum amount that most poor households are willing to sible spend on a latrine. 7. Develop effective compliance monitoring systems The Plan CLTS program promotes far cheaper and simpler INNOVATION: THE GROW-UP-WITH-A-TOILET facilities than the other programs, which should be more PLAN affordable and appropriate for poor households. However, The following plan is proposed to ensure that every child in 35 percent of households in its program communities con- Cambodia “grows up with a toilet� through the provision tinue to practice open defecation, and most of these open of sanitation finance to poor households during the first defecators are likely to be poor households. five years after their first child is born. The intention is that the development of improved sanitation facilities and the The use of public finance to subsidize the development, establishment of good sanitation practices among both par- promotion and marketing of appropriate sanitation prod- ents and the first-born will ensure that the rest of the family ucts is to be encouraged, but there is a risk that the current grows up using a hygienic latrine and observing good sani- sanitation marketing programs will not benefit many poor tation and hygiene practices. viii www.wsp.org Sanitation Finance in Rural Cambodia | Executive Summary The five-year plan would be targeted at poor mothers on the The plan would be supported by demand creation programs birth of their first child, on the basis that poor children un- (CLTS, mass media), sanitation marketing programs to in- der-five are the highest risk group for diarrhea, malnutrition crease and improve the supply of low-cost sanitation goods and worms. Assistance would be provided to the mother of and services, and microfinance programs to enable non- the household to improve household sanitation throughout poor households to develop improved sanitation facilities. the five-year period, with both connection subsidies (incen- tives for the construction of facilities) and outcome-based The intention of the plan is three-fold: firstly, to focus at- sustainability incentives (to encourage long-term improved tention on the need to target sanitation finance towards sanitation practices). improved sanitation among under-five children; secondly, Year 0 (birth of first child) US$15 toilet voucher (redeemable to recognize that sanitation finance should promote a pro- by local producers) cess of sanitation development over a period of several years Plus US$5 rebate on construction of (providing incentives for the upgrading of facilities and the second latrine pit adoption of improved behaviors); and thirdly, to encour- Year 1-5 (annual reward) US$0-10 each year based on age more efficient demand-side financing through vouchers following criteria and cash transfers in place of existing mechanisms for the • Toilet usage (verified) • Village toilet coverage (verified) supply of in-kind materials and services. • Completion of hygiene course • Presence of handwashing facility www.wsp.org ix Abbreviations and Acronyms ADB Asian Development Bank CCT Conditional cash transfer CLTS Community-Led Total Sanitation DHS Demographic and Health Survey ESI Economics of Sanitation Initiative FGD Focus group discussion HH Household IDE International Development Enterprises JMP WHO-UNICEF Joint Monitoring Programme for Water Supply and Sanitation MDG Millennium Development Goal ODF Open defecation free TS-RWSSP Tonle Sap Rural Water Supply and Sanitation Project UNICEF United Nations Children’s Fund USAID United States Agency for International Development US$ United States dollar WASH Water, Sanitation and Hygiene (Sector) WHO World Health Organization WSP Water and Sanitation Program of the World Bank WTO World Toilet Organization x www.wsp.org Glossary of Terms Administrative costs: The costs required to deliver services, including identification of target population, dealing with appeals, processing payments, undertaking monitoring and evaluation, and exercising oversight. Conditional cash transfer: Provision of money to poor families contingent on them making investments in human capital, such as keeping their children in school or taking them to health centers on a regular basis. Error of inclusion: The inclusion of an ineligible person in a program. Error of exclusion: The exclusion of a person who meets eligibility criteria of a program. Externality: A consequence of activity that affects other parties without this effect being reflected in the cost of the goods or services involved. Fee waiver: Exemption from service charges granted to individuals (usually based on personal characteristics such as poverty). Geographic targeting: A targeting method in which location determines eligibility for benefits or allocates budget to concentrate resources on specific areas. In-kind transfer: Provision of additional non-monetary resources to households such as food rations or latrine components. Internalized costs: Costs that are incorporated into the pricing structure of goods or services, such as social costs of the manufacture and use of a product. Market failure: Inefficient production or use of goods and services by the market, often associated with non- competitive markets, externalities or public goods. Means testing: A targeting method based on income that seeks to collect comprehensive information about household income and wealth. Poverty line: Cut-off point separating the poor from the non-poor, either based on monetary (e.g. consumption level) or non-monetary (e.g. a certain level of literacy) measures. Proxy means testing: A targeting method based on fairly easy to observe household characteristics, such as the location and quality of the household’s dwelling, ownership of durable goods, demographic structure, education and so on. www.wsp.org xi Executive Summary Subsidy: Financial measure aimed at controlling the prices of food or other essential commodities and services. Targeting: The effort to focus resources among those most in need of them. Voucher: A printed coupon or ticket that entitles the holder to a discount or can be exchanged for specific goods or services. Also called a “near cash transfer�. xii Economic Assessment of Sanitation Interventions in the Philippines Content 1 Introduction....................................................................................................................................... 1 1.1 Objectives .................................................................................................................................... 2 1.2 Methodology ................................................................................................................................ 2 2 Public Finance of Sanitation ............................................................................................................ 3 2.1 General case for public finance of sanitation ................................................................................ 3 2.2 Development objectives ............................................................................................................... 5 2.3 Targeting ...................................................................................................................................... 6 2.4 Type and level of benefit ............................................................................................................... 9 2.5 Service delivery ........................................................................................................................... 11 3 Conditional Cash Transfers ............................................................................................................. 16 3.1 Relevance for sanitation improvement ......................................................................................... 16 3.2 Features of CCT programs .......................................................................................................... 17 3.3 CCTs for sanitation improvement ................................................................................................ 18 4 Cambodia Case Studies.................................................................................................................. 19 4.1 ADB Tonle Sap RWSSP ............................................................................................................... 19 4.2 Plan Cambodia CLTS program .................................................................................................... 27 4.3 IDE Sanitation Marketing program................................................................................................ 29 4.4 WTO-Lien Aid Sanitation Marketing program ............................................................................... 32 5 Comparative Analysis ...................................................................................................................... 33 5.1 Scale and speed ......................................................................................................................... 33 5.2 Financing mechanisms................................................................................................................ 34 5.3 Cost per household latrine .......................................................................................................... 34 5.4 Public finance efficiency .............................................................................................................. 34 5.5 Investment against income.......................................................................................................... 35 5.6 Operation and maintenance costs............................................................................................... 35 5.7 Summary of comparative analysis ............................................................................................... 36 6 Improved Approaches to Sanitation Finance ................................................................................. 38 6.1 Leave no one out – use a segmented approach .......................................................................... 38 6.2 Check who benefits – monitor targeting effectiveness ................................................................. 39 6.3 Aim for efficiency – recognize market advantages ....................................................................... 39 6.4 Use vouchers to encourage sustainable service provision ........................................................... 39 6.5 Design finance for long-term sanitation practice .......................................................................... 39 6.6 Use national means testing systems ........................................................................................... 40 6.7 Develop compliance monitoring systems .................................................................................... 40 7 Innovation: The Grow-Up-With-A-Toilet Plan ................................................................................. 41 References .............................................................................................................................................. 42 Annex 1 Meetings held in Cambodia .................................................................................................... 44 www.wsp.org xiii I. Introduction This guidance note presents the findings of a study on sani- case studies were supplemented by preliminary data from tation finance in Cambodia conducted for the Water and two recently-launched sanitation marketing projects. The Sanitation Program (WSP). The overall objective of the as- study also examined the potential use and effectiveness of signment was to consider sustainable sanitation financing (hardware) subsidies, conditional cash transfers (CCTs) and options with a focus on promoting access for the poorest. other financing approaches relevant for sanitation improve- ment. This document contains an introduction on sanitation fi- nancing and subsidies, stating the (economic and public The guidance note ends with recommendations for im- health) cases for subsidies as well as some of their practi- proved sanitation finance, including practical suggestions cal pitfalls, in order to introduce and frame the sanitation for sanitation programs in Cambodia. These recommenda- financing debate for a wider audience. The study utilized tions bear particular relevance for the ADB’s Second Rural two case studies of rural sanitation finance in Cambodia Water Supply and Sanitation Sector Project, which com- to illustrate the practical issues, and the data from these menced in 2010. www.wsp.org 1 1.1 OBJECTIVES 1.2 METHODOLOGY The assignment had the following specific objectives: The study involved three weeks spent in Cambodia, includ- • Using existing information to the extent possible, ing field visits to project communities of the ADB Tonle Sap study existing projects and information to develop Rural Water Supply and Sanitation Project (TS-RWSSP); an understanding of the different approaches to consultations with sector stakeholders; attendance at a one- sanitation promotion used in country, and their day World Bank-led seminar on Community Cash Trans- strengths and weaknesses; fers; and secondary research. • Based on available and collected information, ana- lyze one CLTS-based and one subsidy-based sani- Two sanitation finance case studies were prepared, one tation program, using the format and analytical on the ADB TS-RWSSP and the second based on the structure provided through the WSP global study of Plan Cambodia CLTS program. The case studies utilized sanitation financing (thus contributing to the global household survey data collected for the second phase of database on sanitation financing); the WSP Economics of Sanitation Initiative (ESI), nota- • Given an understanding of the pitfalls of traditional bly on household investments in sanitation facilities; and sanitation subsidies (and informed by positive ex- program finance data obtained directly from the ADB and amples in Cambodia and elsewhere), develop alter- Plan program managers. In addition, comparative studies native financing recommendations that avoid issues were made of two sanitation marketing programs by IDE with targeting, administration, technology choice, and WTO-LienAid. perverse incentives, etc.; • Expressly consider the applicability of CCT ap- The case study analysis was based on the template devel- proaches, including practical suggestions for pilot- oped by the WSP Global Sanitation Finance study in or- ing and implementation if such approaches appear der that the findings are comparable with those from other feasible. WSP country studies. 2 www.wsp.org II. Public Finance of Sanitation 2.1 GENERAL CASE FOR PUBLIC FINANCE OF finance has long been used to support the promotion of SANITATION improved sanitation, but with mixed results. The case for the public finance of sanitation rests on the consequence of individual sanitation behavior on the health This study tries to assess whether current approaches to the and well-being of other people. In economic terms, unhy- public finance of sanitation have been effective, and wheth- gienic sanitation behavior by an individual causes a nega- er alternative approaches are likely to result in more socially tive externality, in that bad sanitation practice has adverse or economically optimal outcomes. effects on the health and well-being of others. FEWER PATHOGENS IN THE ENVIRONMENT? In most instances, the polluter does not have to pay for bad Sanitation improvement presents a particular development sanitation practice, thus the practice is more widespread challenge due to the difficulty of generating private demand than it would be if the individual were held accountable for for sanitation facilities. Awareness of the private and exter- these external costs. Furthermore, those who incur these ex- nal costs of inadequate sanitation is generally low in devel- ternal costs do so involuntarily, resulting in outcomes that oping countries. Despite widespread diarrheal disease and are neither socially nor economically optimal. high child mortality rates, health costs are rarely ascribed to unhygienic sanitation practices; toilets are often perceived The market-driven approach to correcting negative exter- to be unaffordable; and demand for improved sanitation nalities is to internalize third party costs by requiring the remains low. polluter to repair or repay any damage caused. However, this is difficult when the true monetary value of the costs is hard to determine, or when the supply and demand for the goods and services are constrained because of market or information failures. Those who pollute the environ- ment through open defecation would be charged the cost of cleaning up the environment and of any illnesses that result from fecal contamination, as well as the cost of the medical services provided to those were ill, and of the lost time and productivity of those affected by the open defeca- tion. Clearly, this would be impractical given the difficulties of attribution, valuation and enforcement. An alternative theoretical approach is for public finance to be used to reduce the negative externalities by increasing the private consumption of the “under-consumed goods�, in this case by using public funds to promote private invest- ment in hygienic sanitation practices. In practice, public www.wsp.org 3 Sanitation Finance in Rural Cambodia | Public Finance of Sanitation Everyone without an improved sanitation facility1 is “sanita- because many sanitation programs require a household con- tion poor� and, therefore, will benefit from public support tribution that is unaffordable to the poorest households. In to improve sanitation. Furthermore, each new improved addition, the rationing and allocation of subsidies through sanitation facility will reduce the number of pathogens in local leaders means that those who are well connected often the environment, thus provide societal as well as private obtain benefits ahead of the poorest households. benefits. In a context of low demand for sanitation, this framework suggests that there is little need for targeting There is a social justice question associated with the use of (among those not using improved sanitation facilities) as scarce resources to provide subsidized sanitation facilities to any new toilet will be beneficial. non-poor households at the expense of poor households. There is also an important issue regarding the alignment Unfortunately, scarce public resources often force policy of sectoral objectives, often focused narrowly on increasing makers to concentrate or ration investments and interven- coverage numbers, with wider national objectives of pov- tions among specific population groups or geographical erty alleviation, welfare improvement and social protection. areas. The best welfare-enhancing approaches would target But, most importantly, there is a risk that the benefits of the poorest first, due to the higher marginal value of each public investment in improved sanitation are limited by an monetary unit among the poorest. But, in practice, sanita- approach that fails to reach the poor, due to the lower dis- tion programs often target the “low-hanging fruit� – those ease and mortality burden found among non-poor house- without improved sanitation facilities who are more willing holds. to invest, more responsive to promotional programs, and easier to reach. The intention of this targeting is that, in Household survey data from Cambodia2 illustrate the ex- addition to the benefits from the additional sanitation fa- tent of the problem. These data confirm that children from cilities, the supply of sanitation goods and services to these poor households have significantly higher mortality and responsive households will build a larger sanitation market, malnutrition rates than those from non-poor households3; thus developing the economies of scale and common good and that the risk of diarrheal disease is three to four times practice that will be needed to change sanitation behavior higher among severely underweight children. and spending priorities among the poorest households. The exact relationship between these health outcomes and LOWER HEALTH AND ECONOMIC BENEFITS sanitation status remains uncertain4, but children from DUE TO INCLUSION ERRORS poor households, particularly those that are malnourished, In Cambodia, where 77 percent of the rural population lack are likely to contribute more pathogens to the environment improved sanitation and 35 percent of rural households are through unsafe excreta disposal than children from non- below the official poverty line, more than half of those with- poor households. Therefore, sanitation strategies that fail to out improved sanitation facilities are non-poor households. deliver improved sanitation to poor households are likely to As a result, sanitation programs that target “low-hanging have less optimal outcomes, and fewer health benefits, than fruit� tend to benefit largely non-poor households. This is those that succeed in reaching the poor. 1 Defined by the WHO-UNICEF Joint Monitoring Programme for Water Supply and Sanitation as “facilities that ensure hygienic separation of human excreta from human contact� in JMP (2008) Progress on drinking water and sanitation: a special focus on sanitation New York: WHO-UNICEF Joint Monitoring Programme for Water Supply and Sanitation. 2 DHS (2006) Cambodia: Demographic and Health Survey 2005 Maryland: ORC Macro and National Institute of Statistics. 3 Under-five mortality rates in Cambodia were found to be 39 percent higher across the bottom two wealth quintiles than in the fourth quintile; similarly, the proportion of severely underweight children was 76 percent higher in the bottom two wealth quintiles than in the fourth quintile. Recent research in Laos indicated that severely underweight children aged 0 to 4 years had 2.8 times higher likelihood of diarrhea than children with normal weight-for-age status, with 3.6 times higher likelihood of diarrhea found in severely underweight children aged 0 to 11 months (analysis based on Lao PDR MICS 2006 household survey data, to be published in UNICEF (forthcoming) Lao PDR: Child well-being and disparities – health, nutrition, water sanitation hygiene, education and protection). 4 Higher sanitation coverage is generally associated with better health outcomes, but the relative importance of the factors that influence these outcomes are not well known; for instance, there is little understanding of the role that unsafe infant excreta disposal plays in determining health outcomes (considering that this practice is often independent of the use of sanitation facilities by adults and older children). 4 www.wsp.org Sanitation Finance in Rural Cambodia | Public Finance of Sanitation Nevertheless, it remains true that excreta-related patho- gens will remain in the environment until everyone uses improved sanitation all of the time7, hence that the gains from investing in collective sanitation outcomes (on a com- munity by community basis) should be higher than those from equivalent incremental increases in sanitation cover- age across a wider population. 2.2 DEVELOPMENT OBJECTIVES There is some tension in the discussion of development objectives between economic development models, which aim to reward entrepreneurship and hard work by putting money into as many pockets as possible, and social devel- opment models, which recognize the market failures that prevent poor and disadvantaged people from playing a full and active part in the economy. The economic development model is related to the “low- IMPORTANCE OF ACHIEVING COLLECTIVE hanging fruit� approach discussed earlier, whereby sup- SANITATION OUTCOMES port to responsive individuals and households is intended In theory, the negative externality associated with unhy- to leverage sanitation investments by others (commercial, gienic sanitation practice means that open defecation or private and non-government) and, eventually, it is hoped unsafe excreta disposal by even one person will adversely that the benefits from these investments will trickle down affect other people. Therefore, the full benefits from pub- to poor and disadvantaged households. lic sanitation finance, which aims to reduce these negative externalities, will not be available unless collective sanita- The social development model suggests that direct inter- tion outcomes, such as open defecation free status or 100 ventions and positive discrimination are necessary to tackle percent improved sanitation coverage within a community, pockets of deprivation, with benefits targeted explicitly at are achieved. poor and disadvantaged households in order to lift them out of poverty, improve their well-being and enable them In practice, the consequence of one person’s, or one house- to be productive members of society. hold’s, unsafe excreta disposal on a wider group proves hard to measure. While some evaluation findings5 suggest Social development approaches are often favored in devel- additional reductions in diarrheal prevalence from open oping countries, but suffer from serious targeting and effi- defecation free (ODF) outcomes, and anecdotal evidence ciency problems. Therefore, any attempt to direct sanitation is mounting of reduced clinical caseloads in rural health finance towards specific segments of the population (poor centers surrounded by ODF communities6, there remain households, mothers, disadvantaged households, people sufficient doubts over the rigor, scale, and cross-sectional living in areas with difficult ground conditions) needs to nature of these findings to question the importance of the utilize a robust targeting system and include explicit mech- external effects of unhygienic sanitation practice. anisms for monitoring the efficiency of service delivery. 5 RMRCT (2007) Quantitative microbial risk based approaches to evaluate Nirmal Gram and Non Nirmal Gram Villages of Rewa District, Madhya Pradesh UNICEF Project report; Knowledge Links (2007) Formative research: development of sanitation IEC manual for Himachal Pradesh. 6 For instance, in Lumajang District, East Java, Indonesia. 7 This broad description includes safe infant and child excreta disposal; use of improved sanitation while away from home (at school, in markets, in the fields); and implies safe disposal of wastes from sanitation facilities. www.wsp.org 5 Sanitation Finance in Rural Cambodia | Public Finance of Sanitation These contrasting development objectives appear to require a mixture of universal interventions designed to stimulate economic development by leveraging sanitation invest- ments from a wide spectrum of the population, and more targeted interventions aimed at providing benefits directly to excluded groups and deprived households. 2.3 TARGETING Concentrating resources on the poor or disadvantaged can increase the benefits that these resources achieve within a given budget, or can achieve a given impact at the lowest cost. In practice, the full theoretical gain is not realized be- cause targeting is never completely accurate, and because costs are associated with targeting8. Programs aim to focus resources on the poor without incur- ring unacceptably high errors of exclusion9 or costs (admin- istrative, private and incentive). Several different targeting methods can often be used for a particular type of program; for instance cash and in-kind transfers can be targeted by means tests, proxy means tests10, nutritional status or risk factors, geographic area, demographic characteristic, or self- selection. For a single program to use a number of methods is common, for example first using geographic targeting to The ID-poor system is now in place, either fully or partially identify poor areas and then proxy means testing to identify over different time horizon, in all provinces in Cambodia, beneficiary households. This combined approach usually and is being used as the principal targeting mechanism yields better targeting than use of a single method11. for Health Equity Funds, which provide health service fee waivers to households identified as poor through their ID- The social protection literature suggests that means tests poor cards. and proxy means tests have the highest costs, but tend to produce the lowest errors of inclusion12 and are often good Implementation of the ID-poor system costs about investments. Self-selection, via setting a low benefit level, US$1.50 per household, with similar investments required and geographic targeting are also powerful and proven tar- every two years to update the system. There are plans for geting tools. third party verification of the ID-poor system in 2010 and, as the number of programs that utilize the ID-poor system MEANS TESTING grows, it seems likely that the accuracy of the targeting and Means testing systems are evolving and improving rapidly. the performance of the system will improve further. Use A new system called ID-Poor has been instituted in Cam- of the same targeting system by several different programs bodia over the last few years, which provides photo-identity should enable the administrative costs to be shared, giving cards to households identified as very poor (Poor Level 1) or the ID-poor system substantial financial advantages over poor (Poor Level 2) based on an asset scoring system. program-specific targeting systems. 8 Grosh et al (2008) For protection and promotion: the design and implementation of effective safety nets Washington DC: The World Bank. 9 Error of exclusion: the exclusion of an eligible person from a program. 10 Proxy means tests use easy to observe household characteristics (such as housing quality, ownership of durable goods, demographic structure and education) as substitutes for measures of income or wealth. 11 Ibid. 12 Error of inclusion: the inclusion of an ineligible person in a program. 6 www.wsp.org Sanitation Finance in Rural Cambodia | Public Finance of Sanitation BOX 1. PROXY MEANS TESTING WHERE ADMINISTRATIVE CAPACITY IS LOW: CAMBODIA’S SCHOLAR- SHIP PROGRAMS Because Cambodia has rather less administrative capacity than the middle-income Latin American countries where proxy means testing originated, it has adapted the general practice of proxy means testing in a way that makes rigor- ous but simplified testing viable. The schools that participate in its scholarship program are subject to a prior round of geographic targeting, and applicants complete a proxy means test that is used to allocate scholarships among each selected school’s students. Cambodia’s CESSP program dispenses with the cadre of field worker/social workers who often administer the in- strument. Instead, students fill out the program application/proxy means test form in school. Then the teacher reads the information aloud and the classmates help verify/certify that it is correct. A local committee of school and com- munity leaders score the forms by hand.a To assist in manual scoring, the formula uses only integers. The ranking is done only within schools, rather than against a national standard as in most proxy means tests. In each school, the scored forms are arranged by score and the poorest children, up to the quota for that school, are selected for the scholarship. This process implies that recipients in poorer schools will be poorer, on average, than recipients in less-poor schools. It is thus less accurate than a ranking against a national standard, but eliminates the need for a national database and the information technology and communications networks that would be required to support it. In a previous scholarship program, the formula was not very sound, so the committees were given leeway to deviate when they thought it appropriate; and when they did so, the students selected were, in fact, poor (as judged later by an evaluation survey). Subsequently the formula was based on statistical analysis of the same type used elsewhere, and the discretion of the local committees was reduced. a In the first year of the CESSP program, an independent firm scored the forms centrally. Source: CESSP scholarship team 2005 in Fiszbein and Schady (2008) Conditional Cash Transfers: reducing present and future poverty Washington DC: The World Bank, Policy Research Report. SELF-TARGETING An examination of these criteria suggests that pour-flush la- Self-targeting goods are those that are widely and dispro- trine pans, which are one of the most commonly and heav- portionately consumed by the poor and are, therefore, most ily subsidized of sanitation goods, are not suitable for self- likely to ensure that any associated subsidies reach the de- targeting. Latrine pans are not inferior goods; consumption sired target population. The best candidates for self-target- of latrine pans is higher by rich households than by poor ing goods are those that meet the following criteria13: households (e.g. positive income elasticity); and latrine pans are tradable. • Universal reach among poor households • Significant percentage of expenditure among poor Recent demand studies in Cambodia15 confirm that there households is strong demand for pour-flush latrine pans, hence that • Inferior goods with low or negative income elastici- many people may be willing to pay for them without sub- ties14 sidy. Given the importance of differentiating self-targeting • Differentiated by quality (e.g. lower quality variants goods by quality, there is a clear case for discontinuing any are used by the poor) subsidy of pour-flush latrine pans in favor of goods that are • Not readily tradable utilized more frequently by poor households. 13 Komives et al (2008) Water, electricity and the poor: who benefits from utility subsidies? Washington DC: The World Bank. 14 The amount consumed (as a percentage of income) rises as income falls. 15 Roberts M (2006) Demand assessment for sanitary latrines in rural and urban areas of Cambodia Phnom Penh: International Development Enterprises; and, Chapin J (2009) Design project: Sanitation marketing pilot project, final report Phnom Penh: IDEO and IDE. www.wsp.org 7 Sanitation Finance in Rural Cambodia | Public Finance of Sanitation INTRA-HOUSEHOLD TARGETING ditions that mandate specific human capital investments, Most sanitation programs target households based on a and transfers that strengthen the mother’s bargaining posi- number of criteria, and then make transfers, or provide tion, will reinforce her ability to shift household spending support, to the “head of the household�. In contrast, con- and time allocation decisions to improve children’s welfare. ditional cash transfers16 (CCTs) are often paid explicitly to mothers, on the basis that mothers have a stronger prefer- AVOID PENALIZING EARLY ADOPTERS ence for investing in children than do fathers17. Many sanitation programs target only households without sanitation facilities, which penalizes poor households that In addition, CCT conditions often apply to use of services are early adopters of improved sanitation, and creates disin- by children aged 0 to 5 years, with the age limit set to al- centives for the self-provision of sanitation facilities. low continuous eligibility from birth to schooling. There is a strong argument for linking sanitation finance to child In contrast, the CCT targeting model provides benefits sanitation outcomes, and for targeting transfers at mothers. to all eligible households, i.e. those that meet poverty and Diarrhea incidence is higher in children than in adults; 95 other targeting criteria, and adopt good behaviors, on the percent of sanitation-related deaths are among under-five understanding that these households (and their communi- children18; and, consequently, many of the pathogens that ties) will require lower public investments in basic services cause serious illness are passed on via unconfined child ex- like health care due to better local outcomes. creta. Sanitation finance should aim to encourage and reward In addition, the inter-generational effects of inadequate good sanitation behavior by poor households and, where sanitation are significant. Children that grow up without possible, avoid disincentives to the early adoption and self- improved sanitation suffer health, nutrition, growth, educa- provision of improved sanitation. The implication is that tion, self-respect and life opportunity deprivations that pre- sanitation finance should seek to promote longer-term be- vent them developing to their full potential. Child mortal- haviors and outcomes rather than just a one-time invest- ity rates are associated with long-term economic growth19, ment in a basic sanitation facility. thus sanitation improvements are likely to have a dramatic economic impact in countries with low sanitation coverage This approach is linked to the political economy argument and high child mortality. of supporting the “deserving poor�. In many countries, in- cluding developed ones, some voters object to the redistri- Many of the serious and irreversible impacts of inadequate bution of wealth through unconditional handouts. The po- sanitation, such as premature death, stunting and impaired litical economy argument suggests that transfers narrowly cognitive development, occur before children are five years targeted at the poor tend to have limited support from the old, thus sanitation investments targeted at pre-school chil- vast majority of taxpayers that finance but do not receive dren are likely to have a significant impact on their well- these benefits, unless these people are “deserving poor� who being and on economic development. are being rewarded for investing in the health or education of their children. This argument reinforces the importance Women generally attach more value to sanitation than do of extending targeting to include all poor households that men, thus are more likely than male household members to invest in improved sanitation and hygiene behavior, wheth- utilize transfers to improve sanitation. In circumstances in er or not these households already own or use an improved which women’s power within the household is limited, con- sanitation facility. 16 Conditional cash transfer: provision of money to poor families contingent on investments in human capital, such as keeping their children in school or taking them to health centers on a regular basis. 17 Fiszbein et al (2009) Conditional cash transfers: reducing present and future poverty Washington DC: The World Bank. 18 Hutton et al (2009) Economics impacts of sanitation in Lao PDR: a five country study conducted in Cambodia, Indonesia, Lao PDR, the Philippines and Vietnam under the Economics of Sanitation Initiative The World Bank, Water and Sanitation Program. 19 WHO (2001) Macroeconomics and Health: Investing in health for economic development Geneva: World Health Organization. 8 www.wsp.org Sanitation Finance in Rural Cambodia | Public Finance of Sanitation COLLECTIVE SANITATION OUTCOMES the transfers, which then further complicates the adminis- Individual targeting criteria for transfer programs, which trative requirements. are usually based on some measure of household poverty or vulnerability, fail to promote collective sanitation out- In-kind transfers should be service specific and aim to boost comes. Sanitation finance mechanisms should also promote under-invested services. In the sanitation sub-sector, most the achievement of collective sanitation outcomes through in-kind transfers consist of construction materials for la- the inclusion of incentives to achieve both improved indi- trines, usually prefabricated items such as concrete slabs, vidual and improved communal sanitation outcomes. concrete rings for lining latrine pits, squatting plates, la- trine pans; or construction materials (cement, sand, gravel, Examples include additional conditions for the payment wood, metal sheets, pipes) for the on-site construction of of benefits, such as in Maharashtra (India), where the pay- household latrines. Few of these items, with the exception ment of post-latrine construction rebates to below poverty of the latrine pans, meet the criteria that they can be used line households is conditional on the verification of open only for sanitation services. defecation free (ODF) status in the community. Another possibility is the payment of additional benefits to poor In-kind transfers are almost always provided up-front, be- households that both use improved sanitation facilities and fore construction or use of the facilities takes place, based live in ODF communities, subject to annual verification of on the understanding that the users are either too poor, or this status. are unwilling, to construct the facilities without the up- front assistance. In many cases, households are asked to dig 2.4 TYPE AND LEVEL OF BENEFIT a latrine pit as a sign of their commitment to the latrine There are essentially two alternative ways of providing di- construction before the in-kind subsidy is provided. The rect assistance to low-income households: cash transfers main condition for the provision of the up-front subsidy and in-kind transfers. is that the facility is constructed, thus there is often little formal monitoring of the use of the facility or of household The social welfare literature suggests that a cash transfer is sanitation behavior after the completion of the facility. the best instrument for addressing concerns about poverty and inequality because it respects the principle of consumer sovereignty, allowing beneficiary households to allocate the additional income they receive to the good or service repre- senting the highest priority for the household20. Cash trans- fers also avoid potential distortions in economic decisions caused by changes in the relative prices of goods. Despite the theoretical advantages of cash transfers, the prevalent instruments of social policy and sanitation fi- nance in most developing countries are in-kind transfers. The practical concerns associated with cash transfers in developing countries, such as administrative constraints linked to limited financial infrastructure, inefficient target- ing mechanisms, and the risk of corruption, result in most programs utilizing in-kind transfers. In addition, there is no guarantee that cash transfers will be used to improve human capital unless specific conditions are attached to the use of 20 Komives et al (2008) Water, electricity and the poor: who benefits from utility subsidies? Washington DC: The World Bank. www.wsp.org 9 Sanitation Finance in Rural Cambodia | Public Finance of Sanitation For instance, a review of the Total Sanitation Campaign The key advantage of an ex post approach – one based on in Andhra Pradesh (India) during 200521 found that while actual results rather than on forecasts of predicted behavior extensive video and photographic evidence confirmed that – is that it is less likely to create market distortions or have the up-front subsidies22 provided by the state government unintended consequences, such as through the finance or had been used to build more than 1.6 million household subsidy of goods and services that prove unpopular or inef- latrines in less than a year, less than 50 percent of the new fective. toilets were actually in use in the villages reviewed. MINIMUM LEVEL OF SERVICE These problems argue for the use of output or outcome Sanitation program finance is often designed around a based finance, whereby rebates or fee waivers are granted to minimum level of service. Conventionally, the minimum households that make specific investments or utilize specific level of service has been defined either by national standards services; cash transfers are made to households that achieve or by an international categorization such as the JMP im- specific outcomes; and service providers are remunerated or proved sanitation facility. The minimum level of service is rewarded based on performance measures such as the num- intended to assure the promotion and use of hygienic sani- ber of new latrines in use, the number of latrine pits emp- tation facilities, thus to ensure that public finance generates tied safely, or the number of low-cost sanitation products improved health outcomes. sold. In reality, the relationships between use of different types Critics of output or outcome based finance suggest that the of sanitation facility and individual or household health administrative costs of the monitoring required to measure outcomes are poorly understood, both because a number sanitation outputs and outcomes exceed the benefits from of other factors (such as water use, hygiene practices, nutri- these approaches; and that output-based aid tends to favor tion, climate, physical conditions, individual behavior and the non-poor, as these households are often better able to collective behavior) influence these outcomes, and because mobilize the resources needed to meet the output or out- there is scant evidence that more expensive sanitation fa- come conditions. In addition, outcome-based approaches cilities provide greater health benefits than basic hygienic risk penalizing people who live in areas where service provi- sanitation facilities23. sion is inadequate. The Disease Control Priorities report24 quotes the World There is little hard evidence to support either viewpoint, Bank Technical Advisory Group conclusion that “the great- as there are very few examples of output or outcome based est determinants of the efficacy of alternative facilities are, sanitation finance in practice. However, there is increasing first, whether they are used by everyone all the time, and recognition that better monitoring of outputs and outcomes second, whether they are adequately maintained. … Pit is required even when up-front subsidies are provided. Giv- latrines would, from the viewpoint of health rather than en undeveloped sanitation markets, it is also critical that convenience, approximate the same rating as a water-borne sanitation programs include a component to improve the sewerage system�. Similarly, a rigorous study of the impact supply of sanitation goods and services, and that any failure of drainage and sewerage on diarrhea in Brazil25 found that to achieve the expected outcomes is carefully assessed to de- “the impact of sanitation in individual households was not termine whether this reflects inadequate demand, ineffec- significantly affected by the type of toilet – there was no tive supply, or other external constraints. significant difference in the incidence of diarrhea between 21 Robinson A (2005) Scaling up rural sanitation in South Asia New Delhi: The World Bank, Water and Sanitation Program South Asia. 22 Provided as a mix of cash and rice vouchers (worth about US$62 in total) to below-poverty line households. 23 Using the JMP definition of an improved sanitation facility as one that hygienically separates human excreta from human contact. 24 Cairncross and Valdmanis (2006) Water supply, sanitation and hygiene promotion Chapter 41 pp.771-792 in Jamison et al (2006) Disease control priorities in developing countries: Second edition Washington DC: The World Bank and Oxford University Press. 25 Moraes et al (2003) Impact of drainage and sewerage on diarrhoea in poor urban areas in Salvador, Brazil Transactions of the Royal Society of Tropical Medicine and Hygiene 97, 153-158. 10 www.wsp.org Sanitation Finance in Rural Cambodia | Public Finance of Sanitation households with a cistern-flush toilet and those with a rudi- mentary pit latrine�. This thinking is reflected in the JMP classification of im- proved sanitation facilities, which includes a wide range of technologies, from simple dry pit latrines to flush latrines connected to sewer systems or septic tanks, all of which are deemed to be hygienic latrines. Public finance should be utilized to increase the use of im- proved sanitation facilities (by everyone all of the time), yet the first step towards improved sanitation is to stop open defecation through a change in individual sanitation behav- ior. Therefore, effective sanitation finance should recognize that sanitation improvement involves a number of different It is also important that the benefit level is set high enough steps over time, including initial behavior change, construc- to have an impact on the beneficiaries’ behavior and invest- tion of a sanitation facility, adoption of improved sanitation ment decisions. Conditional cash transfer (CCT) programs behaviors, improvement of the sanitation facility, and safe typically provide monthly benefit payments that range be- disposal of wastes. tween 1 percent and 29 percent of pre-transfer household expenditures26. In Cambodia, this implies a CCT benefit Therefore, while sanitation policies and strategies should level of between US$1 and US$20 per month, or as much aim for universal use of improved sanitation facilities, sani- as US$240 per household per year27, which is higher than tation finance should be designed to promote and support the transfer provided by most sanitation programs. the adoption of a range of sanitation behaviors, and a range of sanitation technologies that will change as individuals 2.5 SERVICE DELIVERY and communities become more familiar with improved Service delivery systems need to be as efficient and account- sanitation behavior and more willing to invest in improved able as possible in order to reduce administrative costs, pro- sanitation services. Importantly, the setting and financing vide cost-effective services, and avoid losses due to corrup- of any minimum level of sanitation service should not pre- tion or mismanagement. clude or prevent the construction of low-cost hygienic fa- cilities, or constrain the development of innovative, local In general, sanitation programs are implemented through latrine designs. non-private delivery systems, either through public agen- cies (including line ministries, local governments, and pub- MINIMUM BENEFIT LEVEL lic extension workers) or through development partners The administrative cost associated with any benefit transfer (including donor-financed program management teams system – for operation of the targeting, delivery and moni- and non-government organizations). Very few of these or- toring systems – limits the minimum effective size of the ganizations have experience or specialist skills in logistics, benefit. Significant economies of scale are available in the supply chain management, or local procurement systems; administration of national benefit programs, and additional and, even where these skills and experience exist, the rigid economies are available from the use of existing targeting conditions and procurement rules of government and its and financial systems. Nevertheless, the benefit level needs development partners often limit the use of informal service to be set high enough to justify the administrative expen- providers, locally-sourced materials, and local transport and ditures. delivery systems. 26 Fizbein and Schady (2009) Conditional cash transfers: reducing present and future poverty The World Bank, Policy Research Report. 27 Household expenditure data were unavailable, thus these calculations were based on an average per capita consumption by poor households in Cambodia of US$815 per year (adapted from CSES 2007 data, World Bank). www.wsp.org 11 Sanitation Finance in Rural Cambodia | Public Finance of Sanitation Many development programs use competitive bidding pro- to the user of the services (rather than to the financier or cesses to procure local services, but these bidding processes project manager), and therefore to encourage competition are often circumvented or manipulated by contractors who and service improvements among suppliers and service pro- are well connected with those managing the bidding pro- viders. cess. Local service providers often lack the administrative skills to produce bidding documents that conform to the BENEFIT SYSTEMS required standards. The use of more market-based delivery systems for sani- tation improvement implies the provision of purchasing In addition, management of the delivery system by highly power to the target population. The most common forms paid professionals (such as consultants and NGO staff) and of providing this purchasing power are: busy government officials often adds to the expense and complexity of the process. Their remuneration is not usu- • Cash transfers and rebates ally affected by the performance of the delivery system, or • Vouchers and fee waivers the quality of the service provided. In contrast, local sup- • Microfinance ply chains are run by local manufacturers, transporters and retailers, whose pay and profits are determined by market Cash transfers and rebates involve direct payments to target rates, quality of the service provided, and the resultant de- households, usually on verification of meeting any condi- mand for their goods and services. tions attached to the payment, such as a rebate paid to poor households on construction and use of a household latrine. As a result, few development programs are able to provide Ideally, these payments should be made through existing fi- goods or services in low-income communities at prices any- nancial institutions such as banks and microfinance institu- where near as low as those offered by local service providers. tions, which have relatively efficient administrative systems, For instance, recent research by the IDE Cambodia sanita- but this may not be possible in remote rural areas, or among tion marketing project (supported by WSP and USAID) households and individuals that lack formal identification found that local producers were willing to provide the basic or residency papers. below-ground materials for a pour-flush latrine, including ceramic latrine pan, concrete slab, three concrete rings, con- Branchless banking is increasingly popular as a way of deliv- crete pit cover, pipework and transport to nearby villages, ering money and other financial services to people without for a total cost of only US$25. A similar package of latrine bank accounts. Branchless banks use mobile phones, smart materials provided by private contractors through the ADB cards, debit cards and prepaid cards to transmit information Tonle Sap Rural Water Supply and Sanitation Program between the agent, customer and the bank29. Initial efforts (TS-RWSSP) cost an average of US$88 per latrine. While at branchless banking with mobile phones in Cambodia are the IDE project involves the use of innovative design and reported to require the provision of a large number of docu- fabrication techniques to reduce costs, previous latrine cost ments for registration, which is likely to make it difficult for data28 confirm that the ADB delivery system is several times many households; and there remain many poor households more expensive than services available from local providers. that lack easy access to a mobile phone. Therefore, wherever possible, sanitation programs should Cash distributions are often used in developing countries aim to deliver services through local markets rather than that lack reliable financial systems. Cash distributions are through parallel delivery systems. The intention is to reduce based on a list of beneficiaries, each of which presents some the costs of supply, to provide the user with a choice of form of identification, signs the paperwork and receives the service providers, to make the service provider accountable cash. In the Meket Project in Ethiopia, beneficiaries gather 28 Salter D (2009) Sanitation demand and supply in Cambodia: identifying constraints to increasing sanitation coverage Phnom Penh: The World Bank, Water and Sanitation Program Field Note. 29 Grosh M et al (2008) For protection and promotion: the design and implementation of effective safety nets Washington DC: The World Bank. 12 www.wsp.org Sanitation Finance in Rural Cambodia | Public Finance of Sanitation along the road or at the market on the day of payment. There is a risk that a parallel market emerges if beneficiaries Names are called out in groups of ten and the first person resell their vouchers at a discounted value; and service pro- in the group is given the list that all ten people sign, usually viders may attempt to charge a fee to redeem vouchers, or with a fingerprint (to solve the problem of illiteracy and may overprice their goods, if there are any costs associated limit duplicate collections through ink stain checks). The with the provision of the subsidized services. Therefore, it first person on the list also collects the money for everyone is important that the benefit system provides rapid and full in that group, each beneficiary is informed of his or her redemption of any vouchers accepted by local service pro- transfer amount, and the cash is then distributed30. viders, and that some form of monitoring checks that the services purchased with the vouchers are delivered. Vouchers and fee waivers involve program payments to ser- vice providers rather than to beneficiaries. Both approaches MICROFINANCE FOR SANITATION require some form of beneficiary identification (either a IMPROVEMENT printed voucher or an ID card), accreditation and training Micro-lending to poor households is often proposed as a so- of service providers, monitoring and payment systems. The lution to financing sanitation improvements in low-income advantage of these approaches is that they can be imple- communities. Yet there have been few large-scale examples mented alongside non-subsidized services with little disrup- of successful microfinance for sanitation improvement in tion or distortion of the market for services. For instance, rural areas. a voucher scheme could be run in parallel with a sanitation marketing project without undermining the demand for A recent study by the Bill & Melinda Gates Foundation31 services, and can be used to direct customers towards ac- reports that about 60 percent of households who built new credited service providers. toilets under the Total Sanitation Campaign (TSC) in Ma- harashtra took loans with an average size of US$30, which suggests some 640,000 loans totaling US$18 million were utilized for sanitation improvement. The same study reports half a million urban borrowers for sanitation improvements through the Vietnam Bank for Social Policy and the Viet- nam Women’s Union, with an average loan size of US$250 and cumulative loans now exceeding US$100 million. Sanitation-related loans are almost always linked to a wider program of demand creation, and the effectiveness of the demand creation activities appears to be critical to the success of the sanitation microfinance initiative. The Maharashtra and Vietnam examples given above link to well-established programs and institutions with significant outreach capacity. However, the majority of the sanitation loans were made to non-poor rural households, such as in Maharashtra where above-poverty-line households receive no government sup- port for latrine construction, or to low-income urban house- 30 Ibid. 31 Mehta M (2008) Assessing microfinance for water and sanitation: exploring opportunities for sustainable scaling up Bill & Melinda Gates Foundation, Final Report http://www.gatesfoundation.org/learning/Pages/microfinance-for-water-and-sanitation.aspx. www.wsp.org 13 Sanitation Finance in Rural Cambodia | Public Finance of Sanitation holds that have the ability to make regular cash repayments. However, recent evaluations of CLTS interventions in The Vietnam Women’s Union (VWU) provides loans only Cambodia33, India34 and Nigeria35 confirm that the sustain- to households with no other outstanding loan, no previous ability of behavior change, notably the problem of reversion bad debt, and that are deemed able to pay back the loan. to open defecation (even in previously declared open def- Applicants must also have an income above the provincial ecation free communities), remains a challenge. The evalua- poverty line; thus, by definition, they are non-poor. tion findings suggest that CLTS interventions can result in low quality latrines, and that the limited durability of some Despite innovations such as the social collateral approach32 of these latrines risks relatively rapid collapse and abandon- adopted by the Grameen Bank in Bangladesh, few micro- ment, particularly in tropical climates where heavy rainfall, finance institutions are willing to lend to very poor rural high water tables and termite damage present significant households, particularly for non-income generating invest- challenges to low-cost building techniques. ments such as sanitation facilities. Therefore, microfinance is likely to be most effective when used to finance sanitation In addition, the TARU impact assessment suggested that facilities and improvements by non-poor households that incentives for the achievement of open defecation free do not qualify for sanitation subsidies or other sanitation- (ODF) communities, such as the Nirmal Gram Puraskar36 related benefits. (NGP) awards, can lead to short-term interventions, coer- cive approaches and temporary behavior change. DEMAND CREATION Most sanitation programs include a component for sanita- Critics of the CLTS approach suggest that these findings tion and hygiene promotion, which aims to generate de- confirm that higher technical standards are needed; that mand for improved sanitation and encourage the sustained more durable latrines will have lower operation and main- practice of improved sanitation and hygiene behaviors. tenance costs, and higher usage rates, thus will result in bet- ter long-term outcomes. Once again, it is being suggested In the past, this component received a relatively low prior- that, in order to provide better quality latrines to poor ity. The main thrust of sanitation programs, and the bulk of households, some form of initial hardware subsidy will be the finance, was to ensure the construction of new toilets, required. often through the provision of some form of subsidy to in- crease demand for sanitation and enable poor households to This interpretation ignores the reality of India’s Total Sani- afford hygienic facilities. tation Campaign, which already includes hardware subsi- dies and minimum technical standards. The TSC guidelines More recently, approaches such as Community Led Total state that below-poverty-line (BPL) households should be Sanitation (CLTS) have increased the focus on the demand provided with a financial incentive (currently set at INR creation component, arguing that behavior change is the 2,20037, US$45) once they have built and started using an most important and difficult achievement, and that regu- individual household latrine. In several states, these guide- lar and universal use of simple latrines will provide better lines translate into the provision of up-front material pack- outcomes than the implementation of more expensive sub- ages to BPL households. In addition, the TSC guidelines sidized facilities with relatively low usage rates. prohibit the construction of dry pit latrines, hence all new 32 Whereby loans are made to groups, with social capital and peer pressure utilized to minimize defaults as a default by any member of the group prevents the other members from qualifying for further loans. 33 MRD (2009) Community-Led Total Sanitation in Cambodia: Formative evaluation report Ministry of Rural Development, Government of Cambodia, draft report. 34 TARU (2008) Impact assessment of Nirmal Gram Puraskar awarded panchayats: final report UNICEF. 35 Robinson A (2009) Sustainability and equity aspects of total sanitation programmes: a study of recent WaterAid-supported programmes in Nigeria London: WaterAid, Report. 36 Clean Village Award provided to villages in India that meet a set of sanitation-related criteria that include verification of open defecation free status. 37 Government of India, Ministry of Rural Development, Department of Drinking Water Supply Office Memorandum No. W-11037/6/2005-CRSP (Revision of the unit cost of IHHLs under the Total Sanitation Campaign) dated 21 October 2008. 14 www.wsp.org Sanitation Finance in Rural Cambodia | Public Finance of Sanitation latrines are pour-flush water-seal latrines, usually with a concrete slab either provided by the promoter or required to be in place before the incentive payment is made. Despite the technical standards imposed by the TSC, the sustainability of latrine usage and behavior change remains problematic across many parts of India. In most cases, these problems relate to ineffective demand creation rather than to technical shortcomings or inadequate hardware subsidy. Therefore, it is critical that sanitation finance mechanisms are designed to address more than the provision of sanita- tion facilities, as demand creation, supply-side strengthen- ing, capacity development, and outcome monitoring are all central to sustainable and effective sanitation improvement. www.wsp.org 15 III. Conditional Cash Transfers Conditional cash transfers (CCTs) are an increasingly pop- 3.1 RELEVANCE FOR SANITATION ular mechanism for transferring benefits to target popula- IMPROVEMENT tions, generally poor households. A CCT is defined as a Under-nutrition and malnutrition remain significant prob- cash transfer that is conditional on the consumption of a lems in Cambodia, thus the Government of Cambodia has particular good, usually pre-specified investments in the been discussing the introduction of a CCT program for human capital of children. Most CCT programs make reg- nutrition with its development partners. A World Bank- ular payments to poor mothers conditional on the use of supported seminar and workshop was held in Phnom Penh health or education services by her child or children. in October 2009, at which the potential for CCTs was dis- cussed and plans for the development of large-scale pro- Countries have been adopting CCT programs at a prodi- grams were outlined. gious rate38. Virtually every country in Latin America has a program and large-scale programs now operate in Bangla- While a relatively recent phenomenon, the CCT literature39 desh, Indonesia and Turkey, with pilot programs in Cam- contains a number of evaluations that highlight the suc- bodia (CESSP scholarship program, see box earlier), Ma- cess of CCT programs in improving the uptake of health lawi, Morocco, Pakistan and South Africa. and education services such as preventive health checkups, vaccinations, and school enrolment. However, the evidence CCTs have been hailed as a way of reducing inequality; of that these gains result in improvements in final health and helping households break out of a vicious cycle whereby education outcomes is mixed, with some evaluations find- poverty is transmitted from one generation to another; and ing little or no improvement in malnutrition rates and oth- of promoting child health, nutrition and schooling. ers that learning outcomes were unchanged. 38 Fizbein and Schady (2009) Conditional cash transfers: reducing present and future poverty The World Bank, Policy Research Report. 39 Ibid. 16 www.wsp.org Sanitation Finance in Rural Cambodia | Conditional Cash Transfers Recent research has confirmed the long-held suspicion that inadequate sanitation plays a significant role in the nutri- tional status of children. The intuition that diarrheal disease caused by inadequate water supply and sanitation affects nutritional uptake, and that malnutrition in turn increas- es the relative risk of diarrheal disease, is supported by a recent collective expert opinion that about 50 percent of the consequences of malnutrition are caused by inadequate water and sanitation services and poor hygienic practices40. Repeated infections, especially diarrhea and helminthes, caused by poor environmental health lead to underweight (low weight for age) and stunted (low height for age) chil- dren, which, in turn, make these individuals more predis- posed to infections and chronic diseases later in life. Most nutrition programs include child growth monitoring, LEVEL OF BENEFIT supplementary feeding, vaccinations and micronutrient ad- The transfer should compensate households for the op- dition, with CCTs targeted at young mothers with children portunity cost of using the services required by the transfer and paid based on records of service use. Yet few nutrition conditions, subject to overall budget constraints. As noted programs include any components that promote improved earlier, the transfer should be targeted at the poorest for sanitation and hygiene. The rapidly developing field of the best welfare-enhancing solution, and at households that CCT nutrition programs provides an opportunity to ad- cannot support themselves through work (so that the trans- dress the current failure to link sanitation improvement and fers do not discourage the labor supply or crowd out private nutrition, through the potential for additional conditions investments in human capital). that encourage the use of improved sanitation facilities and the achievement of collective sanitation outcomes. CCTs IN PRACTICE CCT systems need to demonstrate that the targeting is 3.2 FEATURES OF CCT PROGRAMS more effective than alternative systems; the delivery system is more efficient; and the outcomes are more beneficial. ARGUMENTS FOR CONDITIONALITY Therefore, most CCT programs involve the use or establish- The attachment of conditions to cash transfers risks compli- ment of a national means testing system; transfer delivery cating the administration of the transfers, hence reducing through banks, mobile phones or reliable local institutions; the efficiency of the delivery system. The main arguments and multiple monitoring and redress systems. for conditionality are that the persistence of irrational be- havior and imperfect information leads to excessive pro- SUPPLY OF SERVICES crastination regarding household investments; conflicts of The CCT literature also recognizes that CCTs alone are not interest within the household (parent to child, and wife to enough to change behavior or improve outcomes. Most husband); and private investment below the socially opti- CCT programs also incorporate promotional and outreach mal level (notably due to unrecognized externalities). In ad- components designed to encourage investments in human dition, government policy is often affected by the political capital; and, in contexts of low capacity and poor gover- economy through lobbying, voting, bureaucracy, and inter- nance, supply strengthening components are needed to agency bargaining, which favor politically and publically make sure that the necessary services are available to the acceptable conditions such as “good behavior�. poor households targeted by CCT programs. 40 Prüss-Üstün and Corvalan (2006) in World Bank (2008) Environmental health and child survival: epidemiology, economics, experiences Washington DC, The World Bank. www.wsp.org 17 Sanitation Finance in Rural Cambodia | Conditional Cash Transfers These elements are a good match for sanitation programs, which generally include sanitation promotion components to build demand and supply-side strengthening to improve the supply and affordability of sanitation goods and ser- vices. One important difference in CCT programs is that the sup- ply of services needs to be carefully monitored to ensure that transfers are not withheld as a result of inadequate service provision (resulting in difficulty in utilizing the services). 3.3 CCTs FOR SANITATION IMPROVEMENT Discussions with social protection experts in Cambodia suggest that there is likely to be little demand or interest in separate CCTs for sanitation improvement, but that there is potential for the inclusion of conditions linked to sanita- tion, perhaps through the inclusion of a top-up payment conditional on sanitation behavior. Given current low sanitation coverage and weak sanitation markets, any CCT sanitation condition would have to be linked to large-scale programs designed to strengthen and improve the supply of sanitation goods and services, and promote improved sanitation behavior. Most health and education CCT programs require moni- toring of service use at the service provider level: at health clinics, and in schools. A key drawback of a sanitation con- dition is the requirement for household-level monitoring, which would greatly increase the number of service points that need to be monitored. However, use of sanitation services could be monitored on a less frequent basis than health or education services, which would partially offset the household-level monitoring requirement. The National Community Empowerment Program (PNPM Mandiri) in Indonesia includes a community cash trans- fer scheme (PNPM Generasi) that provides annual block grants to rural communities, with top-ups conditional on the provision of basic health and education services and the achievement of specific outcome targets. WSP Indonesia has proposed the inclusion of an additional top-up grant conditional on the achievement and verification of open defecation free (ODF) status in the community. This modi- fication is still being debated in Indonesia, but provides a useful model for consideration in Cambodia. 18 www.wsp.org IV. Cambodia Case Studies Four case studies were completed as part of the research gram managers and technical advisers, data provided by the for this study. The first two were detailed case studies of implementing organizations, and analysis conducted by the programs that had been running for three years (as of 2009 author. where the study was undertaken), using preliminary house- hold survey data collected by the WSP Economics of Sani- The template developed for the six-case study WSP sanita- tation Initiative (ESI) Phase 2 study: tion financing study was utilized for the case study analysis in order to allow ready comparability between the Cambo- • ADB Tonle Sap Rural Water Supply and Sanitation dia case studies and other WSP case studies on sanitation Project (ADB TS-RWSSP); finance. • Plan Cambodia CLTS program. 4.1 ADB TONLE SAP RWSSP The other two case studies were partial assessments of sani- Launched in 2006, the US$18 million ADB Tonle Sap tation marketing interventions41: Rural Water Supply and Sanitation Program (TS-RWSSP) aims to provide rural water supply and sanitation facilities • IDE sanitation marketing project; to 1.09 million people in 1,760 villages in the five provinc- • WTO and LienAid sanitation marketing project. es that surround the Tonle Sap lake. Sub-project selection combines elements of poverty targeting with a demand- The information on the case studies was drawn from ESI responsive approach intended to improve the sustainability household surveys (where available), interviews with pro- of project facilities. 41 At the time of the study in late 2009, these sanitation marketing programs were just recently launched. www.wsp.org 19 Sanitation Finance in Rural Cambodia | Cambodia Case Studies TABLE 4.1 ADB TS-RWSSP SANITATION OPTIONS 2005 Cost 2007 Cost No. latrines Latrine type Project subsidy HH cost estimate estimate planned Pour-flush latrine US$50 US$120 40% US$72 79,201 Water seal latrine US$35 US$100 60% US$40 26,110 VIP latrine US$35 US$65 60% US$26 17,407 Dry pit latrine US$20 US$50 90% US$5 24,370 Disabled latrine US$60 US$120 80% US$24 2,031 Total 149,119 HH = Household; VIP = Ventilated Improved Pit The TS-RWSSP is the largest rural sanitation program in ect preparation documents, as latrine costs have increased Cambodia, with an estimated US$5.1 million spent on the dramatically since the project inception. Several revisions sanitation component over the last three years. The project were made to the standard latrine designs utilized by the area includes 3.42 million people, who comprise 25 percent project in a bid to maintain the proposed subsidy levels and of the total population of Cambodia. The project villages minimize costs. In addition, a form of community contract- will include roughly one in four of all villages within the ing was introduced in mid-2009, whereby the provincial five provinces, thus the project aims to make a substantial rural development office calls for bids from local contrac- impact on quality of life and environmental health within tors for each community sub-project, but allows the com- these provinces. munity to select the preferred contractor. The baseline survey conducted at the start of the project Unfortunately, rising unit costs have resulted in the pre- indicated 11 percent sanitation coverage within the proj- mature expenditure of the sanitation budget. The progress ect area, which compares with an average of 20 percent in report (dated 30 September 2009) suggested that 37,115 rural Cambodia42. The project sanitation goal was to build household latrines have been completed, with a further 150,000 household latrines by 2011, which would increase 5,870 contracted but not yet complete (making a total of sanitation coverage across the five provinces by about 21 42,985 contracted latrines). However, expenditure in the percent. “civil works – household latrine� category (31 August 2009) is already at US$2.77 million, exceeding the US$2.38 mil- SANITATION OPTIONS lion budget by 16 percent. The TS-RWSSP promotes four types of household sanita- tion facility: The project team has applied for a major change in project • Pour-flush latrine (water-seal pan with offset latrine scope in order to reallocate funds from other project ar- pit) eas (notably contingencies, consultant fees, civil works for • Water-seal latrine (water-seal pan with direct latrine public latrines and NGO contracts). The majority (61 per- pit) cent) of the US$2.37 million budget adjustment has been • VIP latrine (ventilated improved pit latrine with di- proposed for reallocation to the household latrine budget. rect latrine pit) ADB financial records reveal that the contracted amount • Dry pit latrine (dry pit latrine with direct latrine pit) for household latrines is already at US$3.45 million (al- though only US$2.20 million has been disbursed), thus the The planned project sanitation options and subsidy levels contracted amount is already 45 percent higher than the are set out in Table 4.1. In practice, it proved difficult to original allocation but only 46,000 of the planned 150,000 maintain the theoretical subsidy levels outlined in the proj- latrines will be completed. 42 NIS (2009) Cambodia Socio-Economic Survey: Housing conditions 2007 Phnom Penh: Ministry of Planning, National Institute of Statistics, Report. 20 www.wsp.org Sanitation Finance in Rural Cambodia | Cambodia Case Studies The latrine policy revisions reflect a more pragmatic ap- proach. The original project design involved set percentages for project and household contributions to each household latrine model, which were intended to limit the external subsidy for the more expensive models. In practice, this ap- proach resulted in perverse incentives, for example to re- tain expensive superstructure specifications in order to give the appearance of a lower project subsidy (by proportion), which ultimately raised the level of household contribution beyond that affordable to most rural households. USER CONTRIBUTIONS FOR HOUSEHOLD LATRINES The original project design envisaged that rural households would contribute between 10 percent and 60 percent of It is clear from the financial data that the average cost of the cost of the latrines, depending on the latrine and super- the household latrines has been significantly higher than structure design favored by the household – with the basic expected during the design of the project. The original bud- below ground components of the latrine (concrete rings get for household latrines implied an average “civil works� and latrine slab) provided by the project, and the higher cost (ADB share) of about US$16 per latrine. In practice, user contribution being demanded for more expensive the contracted civil works amount (ADB share) has reached pour-flush latrines with higher quality enclosures. an average of about US$80 per latrine – five times higher than anticipated. In practice, the TS-RWSSP employs a contractor to build the latrines. For pour-flush latrines, the contractor builds In addition, contract data examined during the project im- the entire latrine, leaving the household responsible only plementation unit (PIU) visits in Siem Reap and Kampong for digging the pit and paying the contractor for its share of Thom provinces suggest that the average contracted cost of the latrine cost. During the review visits, beneficiary house- a household latrine reached as high as US$135 – US$140 holds reported that the minimum household contribution in 2009 (using community contracting, which was expected for a pour-flush latrine was about US$80; and that the min- to lower the cost and limit procurement problems), which imum contribution for a dry pit latrine was about US$10 suggests an ADB share of about US$12643. (although some households contributed more in order to build a better latrine enclosure). The TS-RWSSP consultant team reports that significant increases in construction material and fuel prices during Expenditure data collected by the WSP ESI project suggest 2006-08 resulted in a dramatic rise in unit costs. It was also that typical user contributions per ADB pour-flush latrine noted that the standard designs for pour-flush pit latrines are as follows44: (both offset and direct) were modified in 2008 to allow the use of timber or GI sheet for the walls of the latrine en- Cash contribution = US$90 closure, in order to reduce costs from the previous design In-kind materials = US$46 Labor = US$1 wherein all pour-flush latrines had plastered brick walls Total household = US$137 build by the contractor. 43 10 percent of the civil works cost of household latrines is covered by the Government of Cambodia. 44 In-kind materials valued by the respondent households; hours of labor estimated by respondent households and valued based on average rural wage rate of KHR 928 per hour (and a shadow wage rate factor of 0.30). www.wsp.org 21 Sanitation Finance in Rural Cambodia | Cambodia Case Studies suggestion appeared to be borne out by recent community contracts for latrines, which were 100 percent pour-flush latrines. However, it remains unclear whether this message is evi- dence-based, or is a discourse promoted by the project staff and by the community members that have benefited from project assistance45. For instance, 27 percent of all latrines completed by the project have been dry pit latrines, suggesting that there is demand for these simple latrines (perhaps because of the much lower household contribution required). In addition, a larger than average proportion of dry latrines were report- ed in the few villages that have achieved 100 percent sanita- tion coverage. Finally, CLTS interventions across Cambo- dia have demonstrated that many rural people are willing to build simple dry pit latrines, providing the financial cost is relatively low. Questions have been raised regarding the durability of these simple dry latrines, particularly in areas that face high water tables and regular flooding, but there seems clear evidence of initial demand for these low-cost latrines if the household contribution is affordable. No ESI data were available for household contributions to dry pit latrines in ADB project villages, but similar cost data ADB: TARGETING from other programs (in-kind materials and labor) suggest In 2007, data collected by the WSP-IDE sanitation de- that the average household cost for a dry pit latrine (under mand assessment study suggested that 73 percent of rural the ADB program) might be as high as US$48 per latrine households (that do not own a latrine) were willing to pay (US$10 cash + US$37 in-kind + US$1 labor). US$10 for a pit latrine. The same data suggest that just over 40 percent were willing to pay US$30 for a pit latrine (the Interestingly, both villagers and PIU staff reported that cost of the “latrine core� promoted by the IDE sanitation many households that had obtained dry pit latrines from marketing program since 2009); and that only 15 percent the ADB project had subsequently upgraded the latrines to would be willing to pay the US$80 contribution demanded pour-flush latrines (usually costing a minimum of US$10 by the ADB TS-RWSSP. for the pour-flush pan and pipework, with higher upgrade costs if converted to an offset pour-flush latrine). One of the problems reported by the consultants and pro- moters of the TS-RWSSP has been that they have struggled The consistent message from the PIUs, project consultants to create demand for the latrines. Several practitioners not- and village committee members was that villagers prefer ed that demand has developed slowly, with only 20 or 30 pour-flush latrines. It was suggested that dry pit latrines households willing to invest initially, although this number were smelly and unpleasant, and that the wood ash essen- grows as people see the latrines constructed and used, and tial for stopping odors was unavailable in many places. This begin to appreciate their benefits. 45 Cambodia DHS (2000) reported that 98.6 percent of rural households used either firewood, straw or charcoal as cooking fuel, which suggests that wood ash is rarely unavailable in rural areas. In addition, the UNICEF-MRD evaluation of CLTS projects in Cambodia (2009) found that “dry latrines are suitable – when they are used and maintained properly they do not cause bad smell�. 22 www.wsp.org Sanitation Finance in Rural Cambodia | Cambodia Case Studies FIGURE 4.1 WILLINGNESS TO PAY FOR A LATRINE (BY RURAL HOUSEHOLDS THAT DO NOT OWN A LATRINE) $100 $80 Latrine Price $60 $40 $20 $10 73% $0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of Respondents Willing to Pay the Speci�ed Price Source: Roberts and Long (2007) Demand assessment for sanitary latrines in rural and urban areas of Cambodia Phnom Penh: IDE and WSP. The program data discussed earlier suggest that 20 percent materials and skilled labor; and less than 6 percent invested of the households in the ADB project villages have invested in dry pit latrines, despite the much lower entry require- in new latrines, of which 73 percent had built the more ments. expensive pour-flush latrines and 27 percent had built dry latrines. The project expenditure data imply that three- The headcount poverty level in the project area averages quarters of the sanitation investment went on pour-flush about 35 percent, and the ADB design documents for the latrines, all of which were built by non-poor households second phase of the TS-RWSSP note that the selection that could afford the US$80 cash contribution46; while at process for target villages should result in a higher propor- most half of those that built dry latrines are likely to be poor tion of poverty in the selected low-income communities, households (as the investments made in building latrine with perhaps 50 percent of the target village population superstructures indicate that the majority were non-poor being below the poverty line. The IDE willingness-to-pay households). Therefore, only about 10 percent of the sanita- data suggest that these poor households (the bottom third) tion investment has reached those below the poverty line; would be willing to pay around US$12 for a latrine. These the remaining 90 percent has benefited non-poor house- data reinforce the suspicion that the ADB latrine options holds that took advantage of the generous subsidies offered and cost sharing rules required too high a contribution for by the ADB TS-RWSSP. poor households to benefit from the project. Furthermore, these data illustrate that relatively few people Another targeting weakness relates to the location of the were willing to invest in these facilities: only about 15 per- project villages. The maps provided by the PIUs in Siem cent (0.73 x 0.20) of the target population invested in a Reap and Kampong Thom highlighted the clustering of pour-flush latrine, even in the knowledge that they stood the project villages along or close to the main road and its to gain a similar amount from the project contributions in feeder roads. Very few project villages were located in more 46 World Bank adjusted data (based on CSES 2007) suggests US$815 annual consumption by poor households in rural areas of Cambodia. www.wsp.org 23 Sanitation Finance in Rural Cambodia | Cambodia Case Studies remote rural areas, despite awareness that the highest pov- ADB: OPEN DEFECATION erty and lowest service coverage are usually found in remote As noted earlier, open defecation remains a serious prob- rural areas. lem in the project villages. On average, the open defeca- The PIUs noted that, while a number of poverty and WASH TABLE 4.2: ESI HOUSEHOLD SURVEY IN THE ADB TS- RWSSP VILLAGES service criteria are examined in the selection of communi- ties, the determining factor is often accessibility by road, Survey Total No. Survey Household type No. latrines % which is required in order for drilling rigs and contractor’s Pour-flush latrine (program) 82 179 46% vehicles to access the villages to construct water supplies Pour-flush latrine (self-supply) 42 60 70% and deliver materials. Dry latrines - 17 0% ADB: LATRINE USAGE No latrine (program) 76 259 29% Both the review and the ESI household data suggest that Total project households 200 515 39% the use of ADB-subsidized latrines is high in the ADB proj- No latrine (control village) 50 - ect villages. Despite concerns that subsidy-based programs Source: raw survey data provided by EIC (consultant that undertook the household survey) often result in low ownership, limited behavior change, and low latrine usage, the requirement for substantial household contributions appears to have ensured that only households tion rate across all of the project villages was 69 percent, that are genuinely interested in latrine ownership (and can which means that about 748,000 people have no latrine afford the household contribution) have participated in the and regularly practice open defecation (producing 112 tons program. of excreta daily across the 957 project villages). The ESI household survey respondents were selected In Siem Reap province, the ADB monitoring data sug- through a purposive sample, which selected only pour-flush gest that 18 project villages (9 percent) have achieved 100 latrine users and open defecators in the three ADB project percent latrine coverage. The proportion of dry pit latrines villages surveyed. As a result, the ESI survey data are not provided in these villages appeared to be higher than in the representative of the communities from which the house- other villages, confirming findings from other sanitation holds were sampled, and careful analysis is required to inter- projects that the promotion of low-cost latrines is essential pret the relevance of the data regarding program outcomes. for the achievement of ODF communities. In total, 250 households from four villages were surveyed, No data were available on latrine usage and sustainability of which 50 were households practicing open defecation rates in these 100 percent covered communities, but the selected from a non-ADB control village. The 200 house- coverage figures appear a commendable achievement given holds surveyed in the three ADB project villages included starting sanitation coverage of only 2.3 percent. However, 124 households that owned pour-flush latrines, and 76 this achievement needs to be balanced against data showing households without latrines. that 30 of the project villages (16 percent) had achieved less than 10 percent sanitation coverage to date, despite receiv- The households survey data revealed that 42 of the 124 ing some form of water supply intervention and user group households (34 percent) that owned latrines had built them training on improved sanitation and hygiene. without any external assistance. Therefore, only 82 of the 250 households surveyed had latrines built with ADB sup- Given this high level of open defecation, and concerns over port. Among these households, 94 percent reported that the targeting of subsidies at non-poor households, it seems they had “used the latrine yesterday�, and only one latrine- unlikely that the project sanitation investments will have owning household reported that they practiced regular achieved the intended health, economic, social or educa- open defecation. tional benefits. 24 www.wsp.org Sanitation Finance in Rural Cambodia | Cambodia Case Studies ADB: SANITATION PROJECT COSTS A preliminary analysis of the sanitation project costs47 was undertaken to assess the total cost of the latrines provided through the ADB TS-RWSSP. This analysis involved ex- amining the expenditure headings under the project, and allocating the costs to either water supply or sanitation ac- tivities (on the basis of advice from the consultant team). The analysis suggested that the following average costs are linked to the project sanitation development activities: Contracted civil works = US$88 per latrine Software (hygiene promotion, training) = US$16 per latrine Program costs (management, technical assistance) = US$42 per latrine Average project sanitation cost = US$146 per latrine Direct comparisons between the latrine costs in the nascent Household contribution (cash and sanitation marketing programs and those over the last three = US$120 per latrine48 in-kind) years of the ADB TS-RWSSP are difficult, both because the Average latrine cost = US$266 per latrine49 sanitation marketing costs do not include any allowance for the construction of the latrine or for the supply of the This preliminary cost analysis confirms that the TS-RWSSP superstructure materials; and because prices have changed is an expensive service delivery mechanism. The provision significantly over this period50. of program services and contracted civil works through government departments and project implementation Nevertheless, the ADB’s Second Rural Water Supply and units appears more costly than the delivery of these services Sanitation Sector Project proposes that “a sanitation grant through market-based mechanisms, such as through user will be provided to each household, covering the cost of households purchasing similar materials and services from an improved and hygienic dry pit latrine with a concrete private retailers and service providers. ring-lined pit (sub-ground structure only)�51. The proposed project intends to undertake six months of behavior change The sanitation marketing programs currently being imple- and triggering activities in project communities before mented by IDE Cambodia (with support from WSP and providing household sanitation grants of US $75 for the USAID) and LienAid-WTO are promoting the sale of construction of an improved dry pit latrine, based on the US$30 low-cost latrine packages by local concrete produc- following costs: ers, with each package including three concrete rings, a concrete latrine slab, a proprietary concrete pan stand, and Labor to excavate the pit (household) US$15 Materials for the four rings (project) US$44 a ceramic latrine pan. Local producers are willing to waive Materials for the cement slab (project) US$12 the charge for delivery of these packages to nearby villages Construction costs (project) US$13 Supervision costs (project) US$5 (within about 10 km), providing that several orders can be Materials for superstructure (household) US$26 filled in each trip. Total investment US$115 47 In all cases, the sanitation project costs included all costs available from the project information systems, including direct implementation costs (software, hardware, salaries, allowances, program management and support) and indirect support costs (staff training, travel, research and development). 48 Average household contribution in the ESI survey households (including cash, in-kind materials and labor valuation) was estimated to be US$137 for pour-flush latrines; author’s estimate of household contribution for dry pit latrines is US$48. 49 Average cost per “latrine in use� (assuming 94 percent usage) across all latrine types, including cheaper dry pit and VIP latrines. 50 Material prices have risen significantly, thus the average ADB latrine costs are likely to underestimate the current cost of latrine provision through the program. 51 ADB (2009) Proposed Asian Development Fund Grant – Kingdom of Cambodia: Second Rural Water Supply and Sanitation Sector Project: Report and recommendation of the President to the Board of Directors. www.wsp.org 25 Sanitation Finance in Rural Cambodia | Cambodia Case Studies BOX 2. LATRINES COSTS IN CAMBODIA A key element of the cost of household latrines under both the ADB and IDE sanitation programs is the cost of the concrete rings used to line the latrine pit. Material and production costs appear to vary dramatically across Cambo- dia, being lower close to Phnom Penh and international border crossings, and higher in remote rural areas. Recent inflation in material prices also makes it difficult to compare prices with any consistency. The following tables summarize regional cost data on precast concrete rings, with allowance for inflation until 2012 (based on CPI inflation of 7.3 percent in February 2010): TABLE 4.3 COST PER 100-CM PRECAST CONCRETE RING, IN US$ Province 2006 prices (IDE) 2010 prices (ADB) Kandal 3.00 Svay Rieng 4.50 Siem Reap 6.00 10.00 TABLE 4.4 RELATIVE PROGRAM PRICES FOR CONCRETE RINGS, IN US$ Sanitation Program 2006 prices 2009 prices 2012 prices IDE (Svay Rieng improved) - 3.00 3.71 IDE (Svay Rieng normal) 4.50 7.50 9.26 ADB (Siem Reap) 6.00 10.00 12.35 TABLE 4.5 DRY LATRINE COSTS, IN US$ (2009 PRICES) Latrine component IDE core Svay Rieng Siem Reap 4 x concrete rings 12.00 30 40 Concrete slab 1.50 5 8 Total (excluding labor) 13.50 35 48 Total (at 2012 prices) 17.00 43 59 The cost data illustrate the dramatic price inflation that occurred between 2006 and 2009, with recent cost data from Siem Reap suggesting that prices have risen by more than 60 percent over the last 3-4 years. These data also point to large price variations across the country, with the IDE survey in 2006 finding that concrete rings in Siem Reap province were twice as expensive as rings in Kandal province; and the dry latrine cost estimates suggesting that similar components are 37 percent more expensive in Siem Reap than in Svay Rieng province. Finally, the dry latrine cost estimates highlight the significant impact that the reduced concrete production costs predicted by the IDE program could have on overall latrine costs, with the forecast costs amounting to only 40 percent of conventional production costs. The ADB cost estimate suggests that the project latrine cost with four rings) at around US$20, including transport and in the second phase will be US$74 (64 percent), with the 20 percent producer profit. The IDE program has made remaining US$41 provided by the household in labor and efforts to improve the design and reduce the cost of the materials. This project amount includes US$18 for con- latrine materials, and the IDE latrine costs can thus not be struction and supervision costs, thus the total for the latrine taken to represent the current market rate. However, the materials provided by the project is estimated at US$56. spread between the ADB and IDE latrine costs suggest that The IDE sanitation marketing program estimates the cost the ADB project costing is considerably higher than the of the materials for a similar latrine package (dry latrine current market rate. 26 www.wsp.org Sanitation Finance in Rural Cambodia | Cambodia Case Studies UPGRADING DRY PIT LATRINES The ADB’s Second Rural Water Supply and Sanitation Sec- The ADB decision to provide only dry latrine packages in tor Project, proposes to provide dry pit latrine packages, the second phase of the project reflects the learning from builds on this approach by subsidizing a relatively low level the first phase, and appears a positive step towards making of service, thus encouraging self-selection by poor house- the project more pro-poor. The intention is to provide these holds, with an easy and low-cost upgrade route to a pour- dry latrine materials as a starter pack, with support and flush latrine. As noted earlier, there is little need to subsi- promotion provided to encourage those that can afford the dize water-seal latrine pans because demand for these pans upgrade to use these materials to construct pour-flush la- is high and, therefore, those that can afford them are likely trines. to purchase and install them through the market. The design criteria and cost sharing rules adopted in the An important consideration under this approach is the pro- first phase meant that the dry latrine model incurred a proj- cess for emptying or replacing full latrine pits. Most direct ect cost that was similar to that for the pour-flush latrine pit latrines – in which the squatting platform is placed di- model: US$92 average cost for dry pit and VIP latrines rectly above the pit – have relatively low-cost and easy to compared with US$108 average cost for pour-flush and move enclosures, so that the platform and enclosure can water-seal latrines. However, the household costs were very either be moved above a new pit or disassembled while the different: US$10 for the dry pit latrine compared with a pit is emptied. Where the enclosure is more permanent, or minimum of US$80 for the pour-flush latrine. more difficult to move, there is a significant advantage in lo- cating an offset latrine pit so that a second pit can be easily As noted earlier, despite comments by project staff that dry added when the first becomes full, thus avoiding the health latrines were not a good sanitation solution, approximately risks associated with emptying latrine pits that contain fresh 27 percent of the project latrines implemented to date have excreta. been dry pit latrines. The relatively high proportion, given limited promotion by project staff, reflects the greater af- Sanitation programs that encourage a range of latrine op- fordability of the dry pit latrine model. tions and upgrades, and provide significant freedom and choice to the latrine user, must ensure that both the promo- The field visits undertaken in Cambodia confirmed that tion teams and the users understand the importance of de- many of the dry pit latrines implemented by the project, signing for safe pit emptying, or for the addition of a second which required only US$10 cash contribution from the pit, and the potentially costly or harmful implications of household, had subsequently been converted to pour-flush inadequately thought out latrine layouts and construction. latrines through the addition of a water-seal latrine pan and, in some cases, the relocation of the latrine pit in order 4.2 PLAN CAMBODIA CLTS PROGRAM to provide an offset pit (presumably for easier emptying). The Plan Cambodia Community Led Total Sanitation (CLTS) program was initiated as a pilot project in two vil- Given that the conversion from dry pit latrine to pour-flush lages in 2006. Since then, the program has expanded to latrine was reported to cost as little as US$10, this approach cover another 45 villages with a total expenditure of just seems a sensible and rational one: a US$20 cash investment under US$0.5 million during the last two years. then secures a functional pour-flush latrine. The only sig- nificant difference visible between the converted pour-flush The Plan CLTS program targets poor villages with low latrines and the project pour-flush latrines was the quality sanitation coverage and an absence of previous sanitation of the superstructure – the US$80 household contribution projects. The baseline data from the project villages indi- secured a brick built and plastered latrine enclosure, while cate pre-intervention sanitation coverage of only 2 percent. the cheaper converted models used a variety of local ma- The Plan interventions aim to achieve open defecation free terials for the latrine enclosure (corrugated metal sheets, communities where 100 percent of the population use sani- thatch, wood). tation facilities. www.wsp.org 27 Sanitation Finance in Rural Cambodia | Cambodia Case Studies TABLE 4.6: ESI HOUSEHOLD SURVEY IN THE PLAN CLTS VILLAGES52 Survey HH in 3 No. latrines Household type Total survey No. Survey % villages in 3 villages Pour-flush latrine (self-supply) 1 1 1 100% Dry latrines (operational) 80 20 154 13% Dry latrines (non-operational) 85 85 85 100% No latrine 29 29 40 73% Total project households 195 135 280 39% No latrine (control village) 50 - - - Source: raw survey data provided by EIC (consultant that undertook the household survey) Like most CLTS programs, no hardware subsidy is pro- age across the program villages averages 65 percent, which vided to households that build sanitation facilities in the represents a 63 percent increase on the baseline coverage. program villages. The Plan support involves only sanitation and hygiene promotion, community development, capacity The ESI household survey data do not report the propor- building, and monitoring. tion of these latrines that were found to be improved sanita- tion facilities as per the JMP definitions, but it seems likely PLAN: TECHNICAL OPTIONS that some proportion of the simple pit latrines found in The ESI household survey, which covered 245 households these villages were not “easy to clean�, raised above ground (including 50 households from a non-program control vil- level, or supported on all sides54. lage), purposively sampled operational pit latrines, non-op- erational pit latrines and households with no latrine. While Latrine usage is estimated at only 64 percent based on the the total numbers of operational and non-operational la- number of non-operational latrines reported in the ESI trines in the survey villages were unavailable, it appears survey villages. This suggests that about 41 percent of the from the data that almost all of the non-operational latrines project population are using their latrines, with another 23 were surveyed, and only 13 percent of the operational la- percent having abandoned their latrines, and the remaining trines were surveyed. As a result, the overall findings are not 35 percent with no latrines. representative of the community-level sanitation outcomes. The recent MRD CLTS evaluation in Cambodia55 found All but one of the 165 latrine-owning households sur- that a similar proportion of CLTS latrines had been aban- veyed53 had built simple dry pit latrines with wooden or doned during the rainy season due to flooding or collapse. earth-covered platforms directly above unlined latrine pits, However, the MRD evaluation also reported that most of thatched walls and thatched roofs. Eighty-six percent of the the people that had abandoned their latrines were practic- household latrines had been built without external assis- ing “dig and bury� techniques during the rainy season and tance, with only 18 households reporting external or com- planned to return to latrine use once the rains were finished munity support in the construction of their latrine. and they were able to clean and repair their latrines. PLAN: LATRINE USAGE AND OPEN DEFECATION PLAN: LATRINE COSTS The Plan program monitoring data report that at least 16 More than two-thirds of the latrine-owning households (71 of the 47 program villages have been declared ODF, which percent) reported that they had not used any cash in the suggests a 34 percent ODF success rate. Sanitation cover- construction of the latrine. These households estimated that 52 Population data were available for only three of the five survey villages. 53 80 households with operational dry latrines; 85 households with non-operational dry latrines. 54 These criteria are part of the JMP classification of a pit latrine with slab. 55 MRD (2009) Community-Led Total Sanitation in Cambodia: Formative evaluation report Ministry of Rural Development, Government of Cambodia, draft report. 28 www.wsp.org Sanitation Finance in Rural Cambodia | Cambodia Case Studies the local materials they had used in the latrine construction vey data58. Therefore, it is recommended that Plan Cambo- were worth about US$10, and that they had invested an av- dia conduct a thorough survey to establish whether latrine erage of 13 hours of self-supplied labor, which was valued at usage rates are as low as the 64 percent suggested by the US$156. Therefore, the total latrine cost was valued at only ESI data. US$11 including all labor and in-kind materials. 4.3 IDE SANITATION MARKETING PROGRAM The remaining 41 households with self-built toilets report- Implementation of the IDE sanitation marketing program ed an average cash expenditure of US$6 on top of similar began in September 2009 following several months of re- in-kind material and labor contributions, making a total search and development. The program is supported by both cost of US$17. Therefore, the average amount spent on a WSP and USAID with a combined budget of US$760,000 latrine in the Plan households surveyed was less than 10 over a 21-month period. percent of the amount spent on the ADB latrines. The program target is to sell 10,000 household latrines PLAN: PROGRAM COSTS in two provinces using a market-based approach with no A preliminary analysis of the sanitation program costs was sanitation hardware subsidy. The R&D work enabled IDE undertaken to assess the total cost – including program sup- to identify strong demand for a package of latrine com- port and software costs – of the latrines that resulted from ponents that enables rural households to build their own the Plan CLTS program. This analysis involved examining pour-flush latrine, including a basic enclosure, for as little the expenditure headings and amounts in the Plan Cambo- as US$30. While the original intention was to develop a la- dia financial system and allocating the costs to either water trine model at around the US$10 price level, in order to be supply or sanitation activities on the basis of advice from affordable for a substantial number of poor households, the the Plan Water and Environmental Sanitation (WES) ad- human-centered design process led by IDE concludes that viser. few Cambodians are willing to pay for anything less than a pour-flush latrine, and that most villagers would prefer The analysis suggested that the following average costs are to buy an affordable pour-flush latrine at the outset rather linked to the program sanitation development activities: than build a cheaper model and upgrade it in several stages. Software (hygiene promotion, training) = US$48 per latrine Program costs (management, technical assistance) = US$40 per latrine Average project sanitation cost = US$88 per latrine Household contribution (cash and in-kind) = US$19 per latrine Average latrine cost = US$107 per latrine57 This preliminary cost analysis suggests that the Plan CLTS program is an expensive approach, especially when con- trasted with other reports that CLTS software costs in other programs in Cambodia are now as low as US$10 per house- hold. However, it should be noted that the costs reported above are per “latrine in use� thus are increased by the low latrine usage figures estimated from the ESI household sur- 56 Guy Hutton (ESI consultant) advised that a shadow wage labor rate of 30 percent was appropriate in rural Cambodia. 57 Average cost per “latrine in use� (assuming 94 percent usage) across all latrine types, including cheaper dry pit and VIP latrines. 58 Cost (including non-operational latrines) is US$72 per latrine. www.wsp.org 29 Sanitation Finance in Rural Cambodia | Cambodia Case Studies FIGURE 4.2: LATRINE CORE: US$25 LATRINE PACKAGE (EXCLUDING ENCLOSURE COST) LATRINE CORE: US$25 LATRINE PACKAGE (EXCLUDING ENCLOSURE COST) Slab + ceramic pan US$7.20 “Pan stand“ US$1.20 Pit cover slab US$1.20 Concrete ring X 3 US$5.70 Plastic pipe US$1.90 Labor US$1.90 Transport US$0.60 Total cost US$20.50 Profit for producer US$4.50 (22 percent) As a result, IDE invested in the development of a “latrine be able to install the latrine components themselves, which core� that can be readily manufactured by local produc- was one of the reasons for developing a “pan stand� that ers and easily transported to nearby villages. IDE also field enables the local gooseneck latrine pans to be connected tested several different approaches for developing private to the latrine pit without the need for specialist masonry demand for latrine purchases, with the conclusion that skills59. the CLTS approach was the most cost-effective method of generating sanitation demand. The original program bud- The superstructure cost is additional to the US$25 latrine get included a large mass media component for sanitation core. IDE estimates that construction of a simple thatch promotion, but after the field testing IDE decided to uti- enclosure costs only US$5; a larger cement slab and GI cor- lize these funds to hire teams of sanitation promoters that rugated sheeting walls and roof about US$45; and a con- would be used to implement an adapted CLTS approach as crete slab, brick walled enclosure and a GI sheet roof about a precursor to the latrine marketing effort made by trained US$125. local producers. The IDE willingness-to-pay data suggest that less than 43 IDE also developed higher cost options including an percent of households without a latrine will be willing to upgrade package to convert the latrine core into a twin pay US$30 or more for a new latrine. Given that 50 percent pit latrine, which will cost only US$14; and a tiled slab of those without latrines are below the poverty line, these latrine core package with four concrete rings that costs data imply that poor households will not be able to afford US$30. The IDE approach assumes that purchasers will the IDE sanitation core 59 Standard pour-flush latrine installations in Cambodia require the construction of a brick and cement chamber under the latrine pan containing a sloping channel that directs the wastes into a pipe connected to the latrine pit. 30 www.wsp.org Sanitation Finance in Rural Cambodia | Cambodia Case Studies FIGURE 4.3: WILLINGNESS TO PAY FOR A LATRINE (BY RURAL HOUSEHOLDS WITHOUT A LATRINE) $100 ADB $80 Latrine Price $60 50 percent of those without toilets are poor (max. WTP US$22) $40 IDE $30 $20 $10 15% 43% 73% $0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of Respondents Willing to Pay the Speci�ed Price IDE has made some efforts to develop lower costs latrine pendent on bulk orders from each village, as producers will models, including a dry pit latrine and a bamboo-lined la- not be willing to transport goods for only one household. trine pit, but it is clear that the bulk of the program resourc- es and activities are linked to the more expensive US$30 Furthermore, the IDE program is not aiming for 100 per- sanitation core. cent coverage in its target villages. The main program objec- tive is to sell 10,000 improved sanitation facilities, with no However, IDE reports that its initial village marketing cam- incentives for achieving collective outcomes, such as ODF paigns have generated sales right across communities, with status. even poor households committing to buy the US$30 latrine core. Willingness-to-pay estimates are notoriously unreli- IDE: PROGRAM COSTS able, given the difficulty of averaging highly variable and While no expenditure data are yet available from the IDE easily influenced spending intentions across large and di- program, the program budget was used to estimate the verse populations, thus it is possible that effective market- planned cost per latrine. Given the market-based nature of ing of a more desirable sanitation product has shifted the the program, whereby households make decisions to invest willingness to pay curve upwards, thereby capturing a far their own resources in sanitation facilities, it is hoped that higher proportion of those without latrines at the US$30 latrine usage rates will be relatively high. price point. The cost figures presented below make no allowance for IDE: TARGETING non-usage, collapse or breakage of the IDE-supported The IDE program will target areas within 20 km of main household latrines, thus may be underestimates of the unit roads in order to reduce the delivery costs for its producers, cost per “latrine in use�. The cost figures are based on the thus will exclude remote villages with higher levels of pov- program goal of selling 10,000 latrine packages in only erty and lower sanitation coverage. The program is also de- 21 months, which may be exceeded (if the program is ex- www.wsp.org 31 Sanitation Finance in Rural Cambodia | Cambodia Case Studies tremely successful). But the short timeframe suggests that The WTO-LienAid program was not yet in the full imple- it is more likely that this figure is ambitious, in which case mentation phase, but it planned to sell 4,000 latrine pack- the costs presented below will be underestimates of the true ages over the following twelve months. Given a total budget program cost per latrine. of US$338,000, achievement of this target will mean a pro- gram cost per latrine of about US$84. The WTO-LienAid The program cost data suggest that the following average intention to market latrine enclosures as well as the latrine costs will be linked to the program sanitation development core suggests that the average household contribution is activities: likely to be higher than in the IDE program. Software (hygiene promotion, training) = US$20 per latrine Program costs (management, The program cost data suggest that the following average technical assistance) = US$56 per latrine costs will be linked to the program sanitation activities: Average project sanitation cost = US$76 per latrine Software (marketing, research, strategy) = US$28 per latrine Household contribution (cash and Program costs (management, technical in-kind) = US$60 per latrine assistance) = US$56 per latrine Average latrine cost = US$136 per latrine Average project sanitation cost = US$84 per latrine Household contribution (cash and in- The total cost per latrine is greater than the US$107 per kind) = US$80 per latrine latrine estimated for the Plan CLTS program due to the Average latrine cost = US$164 per latrine higher estimation of household investment. However, the IDE costs include only US$20 per latrine for the CLTS promotion and other marketing activities, compared with US$48 for software activities in the Plan program; and the IDE program expects to provide a far higher level of service – a pour-flush latrine with a lined latrine pit – than that provided under the Plan program. 4.4 WTO-LIEN AID SANITATION MARKETING PROGRAM The World Toilet Organization (WTO) and LienAid sani- tation marketing program builds on the research and de- velopment work done by the IDE program, supplemented by its own research on reasons for investment (and non- investment) in sanitation facilities. One of the key differences between the two sanitation mar- keting programs is that the WTO-LienAid program has identified the provision of the superstructure as critical to the success of sanitation interventions. Their research sug- gests that one of the factors influencing low latrine usage rates is the failure to complete a superstructure, which then limits the comfort and privacy of the facility. Therefore, at the time of the study, the WTO-LienAid program was in the process of developing a low-cost and mobile “flat- pack� latrine enclosure that it planned to market for around US$50. 32 www.wsp.org V. Comparative Analysis The comparative analysis examines the four programs The following charts present a comparative analysis across (ADB, Plan, IDE, WTO-LienAid) using a set of param- the four case studies. However, it should be noted that the eters developed by the WSP sanitation financing study. The data included for the IDE and WTO-LA sanitation mar- following table summarizes the key variables that were used keting programs are based on planned outputs rather than to produce the charts shown in the sections below. actual outcomes. As noted earlier, some of the marketing targets set by the programs are ambitious, given relatively TABLE 5.1 COMPARATIVE ANALYSIS OF SANITATION short intervention periods, hence it is possible that some of FINANCE PARAMETERS the data presented below will prove to be overestimates of ADB Plan IDE WTO-LA the eventual performance of the marketing programs. Number of people 163,974 12,149 48,000 19,200 served Facilities built per year 13,955 1,321 5,714 4,000 5.1 SCALE AND SPEED Hardware subsidy per US$88 US$0 US$0 US$0 While all of the programs serve (or plan to serve) more than household 10,000 people, the ADB program is several times larger Hardware subsidy (% 42% 0% 0% 0% than the others. The programs are also of similar duration, of cost) thus the pattern formed by the number of sanitation fa- Hardware subsidy per US$88 US$0 US$0 US$0 cilities built (or planned) per year is similar to the program unit scale pattern. Software support per US$59 US$89 US$76 US$84 unit FIGURE 5.1 PROGRAM SCALE Household investment US$120 US$18 US$60 US$80 per unit Total cost per unit US$266 US$107 US$136 US$164 200,000 Leverage ratio 82% 20% 79% 95% (household to program) Number of people served Increased access 7 11 13 12 per US$1,000 public 150,000 finance Hardware cost (% 14% 1% 4% 5% average income) Hardware cost (% 26% 2% 7% 10% 100,000 poor income) Household investment 8% 1% 4% 5% (% average income) Household investment 15% 2% 7% 10% (% poor income) 50,000 Inclusion error 90% - - - Exclusion error 97% - - - Non-subsidized 58% 100% 100% 100% LienAid investment (% total 0 Plan IDE ADB invest) www.wsp.org 33 Sanitation Finance in Rural Cambodia | Comparative Analysis FIGURE 5.2 PROGRAM SPEED 5.3 COST PER HOUSEHOLD LATRINE The Plan program achieves the lowest cost per latrine due 15,000 to the significantly lower household contributions involved. FIGURE 5.4 TOTAL LATRINE COSTS 12,000 $300 Facilities built per year $250 Cost per household latrine 9,000 $200 $150 6,000 $100 3,000 $50 LienAid $0 IDE Plan ADB Household investment per unit IDE LienAid Software support per unit 0 Plan ADB Hardware subsidy per unit 5.2 FINANCING MECHANISMS 5.4 PUBLIC FINANCE EFFICIENCY Figure 5.3 confirms that the ADB program is the only one The Plan program has the lowest leverage ratio of the four of the four that provides a sanitation hardware subsidy programs. The other three programs achieve leverage ratios FIGURE 5.3 FINANCING MECHANISMS 100% close to parity, which means that every dollar invested in the program leverages an equivalent spend by the user house- $100 hold, whereas the Plan program spends US$5 per US$1 invested by the household. 80% Hardware subsidy as percent of total cost FIGURE 5.5 HOUSEHOLD INVESTMENT PER DOLLAR OF $80 PROGRAM EXPENDITURE 60% Leverage ratio: US$ household investment per 1.0 Hardware subsidy in US$ $60 US$ program expenditure 40% 0.8 $40 0.6 20% 0.4 $20 0% 0.2 IDE LienAid LienAid $0 Plan 0.0 IDE ADB Plan ADB Hardware subsidy per household Leverage ratio Hardware subsidy as percent of total cost 34 www.wsp.org Sanitation Finance in Rural Cambodia | Comparative Analysis FIGURE 5.6 NUMBER OF SANITATION FACILITIES PER In most countries, expenditures of less than 5 percent of US$1,000 PUBLIC FINANCE annual income on water and sanitation services are deemed affordable. No data were available on water supply costs, 15 but the financial data suggest that the Plan program is af- Bang for buck: facilities per US$1,000 fordable, requiring household investments of only 2 per- 12 cent of the annual consumption of a poor household. The other three programs all require households invest- 9 ments higher than 5 percent: 7-10 percent in the IDE and LienAid projects, and 15 percent in the ADB program. These data reinforce the earlier findings that most of the 6 sanitation facilities supported by these programs are likely to be constructed by non-poor households. 3 5.6 OPERATION AND MAINTENANCE COSTS LienAid 0 IDE The ESI household survey provided some information Plan ADB Increase access about operation and maintenance (O&M) costs. In the per 1,000 US$ public �nance Plan CLTS program, 80 percent of the dry pit latrine users reported no repairs and no O&M costs, while the remain- 5.5 INVESTMENT AGAINST INCOME ing 20 percent made repairs to the latrine enclosure that The Cambodia poverty data obtained from the World were valued by the users at zero cost. In the ADB program, Bank suggested low annual consumption levels: US$1,525 96 percent of the pour-flush latrine users reported no re- per year average consumption for rural households, and pairs and no O&M costs, while the three households that US$815 per year average consumption for poor rural made repairs to their latrine enclosures spent an average of households. US$45 per latrine. FIGURE 5.7 SANITATION INVESTMENT COST AGAINST ANNUAL HOUSEHOLD INCOME 30% 25% 2%0 15% 10% 5% LienAid IDE Plan Hardware cost as % average income 0% ADB Hardware cost as % poor income Household investment as % average income Household investment as % poor income www.wsp.org 35 Sanitation Finance in Rural Cambodia | Comparative Analysis At the time of the study, none of the latrines in the ADB many years, and soap is cheap. The major maintenance cost or Plan programs is more than two years old, and no full relates to emptying a full pit, or to the relocation of the la- latrine pits or pit emptying costs were reported. trine above a new pit. The Bangladesh study estimated that the pit emptying cost averaged about US$1 per year, even The limited O&M analysis highlights the fact that, while when paid to others. low-cost CLTS toilets may require more frequent repairs than the more durable pour-flush designs implemented un- The conclusion of this study is that careful valuation of der the ADB program, these repairs usually cost very little. operation and maintenance costs is required to provide When the average cost of the latrine is only US$13, and the a realistic picture of the impact of these costs. While it is majority of this amount is the value of local materials used true that regular repair and rebuilding costs are likely to in the construction (e.g. thatch for the walls and roof ), then be a disincentive to sustainable latrine usage, if these costs most repairs are likely to be affordable. In contrast, more are nominal (e.g. valued by the households at zero cost), expensive latrines with rendered brick enclosures are likely then this is unlikely to be a major factor in latrine use prac- to last longer without repair, but any repair costs are likely tices. to be higher and require cash expenditures for the purchase of cement and other market-bought materials. 5.7 SUMMARY OF COMPARATIVE ANALYSIS The comparative analysis confirms that public finance for The WSP sanitation financing study60 concluded that sanitation in Cambodia is not reaching those below the where operating costs were high as a proportion of invest- poverty line. Ninety percent of the public finance for the ment costs – such as in the Bangladesh program, where large ADB program goes to non-poor households, and the annual operating costs were estimated to be 30 percent of two sanitation marketing programs will require households the latrine investment cost – “savings may be achieved by to contribute at least US$30 in order to obtain a latrine, building more solid latrines in the first place�. This conclu- whereas the willingness to pay data imply that US$10 is the sion assumes that the annual operating costs are a burden maximum amount that most poor households are willing to on the household, but neglects the fact that these operating spend on a latrine. costs are actually minimal because of the low initial cost of the latrine. Despite the assumed burden, the case study also The Plan CLTS program promotes far cheaper and simpler noted that most latrines were clean and well maintained, facilities than the other programs, which should be more which implies that few households have problems in under- affordable and appropriate for poor households. However, taking the regular operation and maintenance tasks. the latrine coverage and usage data suggest that, on aver- age, 35 percent of households in the program communities Significantly, the Bangladesh operating costs were overesti- continue to practice open defecation. Given progressively mated. It was assumed that the pit would be emptied every lower sanitation coverage among the poor, it seems likely year, which would be unusual; that all households would that the bulk of those not reached by the program – those pay someone to empty their pit, or to dig a new pit and not, or no longer, practicing improved sanitation – are poor relocate the latrine, whereas in practice many poor house- households. holds would undertake this work themselves; and that cash purchases for soap, sandals, brooms and water pots would Demand creation is critical to all of the sanitation programs. amount to US$4 per year. The ADB program implementers report that significant efforts were required to generate demand for the latrines In practice, as found by the ESI survey of Plan program promoted, despite the substantial subsidy offered by the households, poor households make very few purchases for program. Furthermore, the promotion of highly-subsidized the operation and maintenance of latrines. Brooms are pour-flush latrines by previous programs has created high made from local materials, plastic water pots are used for expectations among low-income communities. 60 Tremolet S, Perez E and Kolsky P (2009) Financing on-site sanitation for the poor: a global six country comparative review and analysis The World Bank, Water and Sanitation Program, draft report. 36 www.wsp.org Sanitation Finance in Rural Cambodia | Comparative Analysis The IDE sanitation marketing program found that an is directed towards developing and marketing products and adapted form of the CLTS approach was the most cost- services that are specifically targeted at the poorest house- effective method of generating demand for its sanitation holds and those that cannot afford the US$30 sanitation products. It was also noted that communities where CLTS core package. had already been implemented exhibited much stronger de- mand for a range of sanitation products, including low-cost In addition, most of the sanitation programs promote sin- dry latrine products, than other communities. gle pit latrines, which require emptying when full or, if a direct pit latrine, removal and relocation of the platform Sanitation programs need to promote sanitation options and latrine enclosure. No sanitation finance is provided to that are affordable and appropriate to poor households, and support this critical process, and little monitoring is carried also need to address sustainability issues. The main ADB out, thus there is a substantial risk that the pit emptying sanitation model required a minimum US$80 cash contri- results in unsafe disposal of fresh excreta within or around bution, which meant that only non-poor households were the village, or that latrine usage is abandoned when the la- able to gain the US$107 subsidy offered by the program. trine pit is full. The high program cost of this approach also limited the number of latrines that could be subsidized, thus resulting Finally, few of the programs have been successful in achiev- in a much lower increase in sanitation coverage than ex- ing collective sanitation outcomes, which should be the ul- pected. timate aim of all sanitation programs (in order to eliminate externalities). The population segment that practices open However, the initial findings of the IDE sanitation mar- defecation in the program communities is largely made up keting program suggest that willingness-to-pay for more of poor households, and generally includes those with the desirable and targeted sanitation products may be higher highest disease burdens, hence those that are most likely to than previously estimated. Nonetheless, many of the poor- transmit diseases to others through unsafe excreta disposal. est households – those with the most acute health, social As a result, the benefits achieved by these sanitation pro- and economic burdens – have limited cash availability, thus grams are likely to be limited. are unlikely to purchase market-bought products unless the prices are significantly lower. The use of public finance to subsidize the development, promotion and marketing of appropriate sanitation prod- ucts is to be encouraged, but it appears that the current sanitation marketing programs are unlikely to benefit many poor households. Given that each IDE latrine package cur- rently costs US$76 in software and program support costs – more than double the cost of the latrine core – a substan- tial transfer is being made to non-poor households. It is hoped that these investments will contribute to the devel- opment of sustainable local supply chains, which should enable long-term improvements in the availability, cost and quality of sanitation goods and services. In addition, it is likely that the marketing cost per latrine package will de- crease over time as the initial development is completed and understanding of market drivers improves. Nevertheless, it is important that an appropriate amount of public finance www.wsp.org 37 VI. Improved Approaches to Sanitation Finance 6.1 LEAVE NO ONE OUT – USE A SEGMENTED APPROACH does not imply that everyone has to adopt identical policies The lessons from the case studies and theoretical analyses but that policy clashes should be identified and, wherever suggest that a segmented approach is needed, with comple- possible, modified to improve the complementarity of the mentary programs designed to ensure that each of the un- programs. served groups (or market segments) receives some form of assistance in developing improved sanitation facilities and The IDE sanitation marketing program provides an exam- adopting improved sanitation and hygiene behaviors. ple. An NGO working in the IDE project area was pro- viding free US$300 latrines to selected poor households. A segmented approach implies not that every sanitation IDE was concerned that these subsidized latrines would program has to target all population groups, but that the lessen demand for its non-subsidized sanitation cores, thus sanitation sector adopts a more harmonized and coordi- approached the NGO to discuss the matter. As a result of nated approach, whereby the target group for each program these discussions, the NGO agreed to revise its approach to is clearly identified, and any gaps are highlighted so that focus on the provision of subsidized above-ground latrine complementary programs can be added. enclosures to poor households that had already bought the IDE latrine core. This solution allowed the NGO to contin- This sectoral approach also requires that individual sanita- ue to provide assistance to poor households, expanded the tion programs do not undermine each other though, for ex- number of households that the NGO could afford to assist, ample, by adopting policies that are likely to lessen demand and provided additional incentives for poor households to for other sanitation services. A more coordinated approach buy the IDE latrine cores: a win-win outcome. 38 www.wsp.org Sanitation Finance in Rural Cambodia | Improved Approaches to Sanitation Finance Social protection literature suggests that direct support far more efficient and cost-effective at delivering sanitation should be provided only to the poorest. In Cambodia, the goods than local government systems, even if (as in the most easily identifiable group for direct support are food- ADB program) service delivery is contracted out to private poor households, which are loosely classified as Poor Level service providers. 1 by the ID-Poor system. The remainder of those without sanitation include poor households above the food poverty Competition should be encouraged by ensuring that mul- line, and non-poor households. Both groups would benefit tiple service providers are available in each locality, and that from the promotion of improved sanitation and hygiene, transfer beneficiaries are free to choose locally accountable and from an improved supply of affordable and appropri- service providers. ate sanitation goods and services. It also recommends that microfinance programs should be developed to provide 6.4 USE VOUCHERS TO ENCOURAGE loans to non-poor households for sanitation improvement, SUSTAINABLE SERVICE PROVISION perhaps as part of broader finance packages for home and The advantages of a market-based delivery system argue for environmental health improvement. transfers that allow beneficiaries to select service providers based on reputation, price and preference, rather than pro- 6.2 CHECK WHO BENEFITS – MONITOR gram-driven decisions regarding the most efficient or effec- TARGETING EFFECTIVENESS tive provider. This approach should help to develop sustain- At present, few sanitation programs are effective in reaching able local supply chains that continue to provide services poor households, thus greater and more specific efforts are and develop new products even after sanitation programs needed to target benefits more closely, and to monitor the are finished. effectiveness of this targeting through the measurement of inclusion and exclusion errors. This finding suggests that demand-side transfers, such as vouchers for latrine materials and rebates for latrine con- The inclusion of a relative wealth ranking – usually con- struction, may be the most efficient form of sanitation fi- ducted through a short participatory assessment – in the nance. Cash vouchers redeemable at local producers could baseline activities of all sanitation programs would allow be linked to sanitation marketing programs, replacing cur- an easy assessment of post-intervention targeting outcomes. rent systems designed to deliver specific sanitation goods to Specifically, sanitation programs should measure the num- rural households. ber and proportion of poor households that are reached by their program, and the number of non-poor households A fixed value voucher could be linked to a minimum level that benefit. In addition, feedback loops are required so of sanitation service, with provision for some contribution that targeting and service delivery systems can be improved by the household. Eligibility for a latrine voucher could be when poor targeting performance is detected. linked to existing means testing systems such as the ID- poor, with additional criteria, such as households contain- 6.3 AIM FOR EFFICIENCY – RECOGNIZE ing children under five, added to reduce the number of MARKET ADVANTAGES beneficiaries where resources are limited. The case study analysis highlights the relative efficiencies of the different delivery mechanisms. Despite similar expendi- 6.5 DESIGN FINANCE FOR LONG-TERM tures on software and program support, the latrine materials SANITATION PRACTICE delivered by the ADB program cost more than double those The focus of most sanitation programs is on the short-term; provided through the IDE sanitation marketing program. on promoting or providing support for the construction of a sanitation facility. Few sanitation programs contain provi- While a more detailed study is needed to identify the details sion for assistance for upgrading latrines over time, for the and modalities of the different cost advantages, it appears construction of a second latrine pit, or for the provision of that market-based mechanisms – whereby consumers pur- benefits to households that maintain good sanitation and chase goods and services from private service providers – are hygiene practices. www.wsp.org 39 Sanitation Finance in Rural Cambodia | Improved Approaches to Sanitation Finance An improved approach to sanitation finance should recog- Additional demographic and geographic requirements nize that the initial construction of a household latrine is could be added to the means testing system, such as tar- only the first step in a process towards improved sanitation geting mothers with children aged under five who live in and the embedding of good sanitation habits. The CCT households classified as Poor Level 1 in areas with high approach suggests that sanitation finance should be used to child malnutrition rates. provide incentives for improved behavior, for increased use of sanitation services, and for improved sanitation among 6.7 DEVELOP COMPLIANCE MONITORING children. Specific mechanisms should be designed to fi- SYSTEMS nance safe disposal of full latrine pits and development of a Improved monitoring is central to sanitation development, second latrine pit. and remains one of the main challenges of sanitation pro- grams. Improved approaches should encourage household Few existing sanitation finance mechanisms contain any WASH cards that record household facilities, monitor facil- provision for finance dependent on children’s sanitation ity condition, register construction of a second latrine pit, behavior, or for monitoring of child sanitation outcomes and provide usage records that can be monitored by transfer as a measure of program performance. The effects of early schemes. The same cards could also be used to record at- deprivation, such as stunting, have lasting impacts on child tendance at hygiene promotion sessions and completion of development, while the habits and practices learned dur- hygiene improvement courses, using a similar model to the ing these early years are often retained throughout life. An Community Health Clubs61 developed in Africa. improved approach should target children specifically and should provide specific incentives for the development of Local government bodies, NGOs and CBOs should be improved sanitation behavior and outcomes during the first used to verify targeting and monitoring systems, alongside five years of a child’s life. social accountability tools such as citizen report cards and score cards that hold service providers accountable. 6.6 USE NATIONAL MEANS TESTING SYSTEMS Sanitation programs often rely on their own targeting sys- tems. For a number of reasons, these targeting systems are rarely effective in directing benefits to poor households. The case studies in Cambodia confirm the targeting problems faced by programs reliant on self-targeting or on targeting criteria developed for combined water supply and sanita- tion interventions. Wherever possible, sanitation programs should aim to base targeting mechanisms on existing large-scale means testing systems, such as the ID-Poor system in Cambodia. These systems use objective criteria and usually benefit from econ- omies of scale that reduce administrative costs. Most target- ing systems have some shortcomings, but national systems that are used to target social protection programs are likely to be more reliable for targeting benefits to poor and vul- nerable households than most local systems. 61 Waterkeyn J and Waterkeyn A (2005) Taking PHAST the extra mile through Community Health Clubs: the AHEAD methodology Africa AHEAD website. 40 www.wsp.org VII. Innovation: The Grow-Up-With-A-Toilet Plan The following plan is proposed to ensure that every child in The intention of the plan is three-fold: firstly, to focus at- Cambodia “grows up with a toilet� through the provision tention on the need to target sanitation finance towards of sanitation finance to poor households during the first improved sanitation among under-five children; secondly, five years after their first child is born. The intention is that to recognize that sanitation finance should promote a pro- the development of improved sanitation facilities and the cess of sanitation development over a period of several years establishment of good sanitation practices among both par- (providing incentives for the upgrading of facilities and the ents and the first-born will ensure that the rest of the family adoption of improved behaviors); and thirdly, to encour- grows up using a hygienic latrine and observing good sani- age more efficient demand-side financing through vouchers tation and hygiene practices. and cash transfers in place of existing mechanisms for the supply of in-kind materials and services. The five-year plan would be targeted at poor mothers on the birth of their first child, on the basis that poor children un- der-five are the highest risk group for diarrhea, malnutrition and worms. Assistance would be provided to the mother of the household to improve household sanitation throughout the five-year period, with both connection subsidies (de- signed to provide incentives for the construction of facili- ties) and outcome-based sustainability incentives (designed to encourage long-term improved sanitation practices). Year 0 (birth of first child) US$15 toilet voucher (redeemable at local producers) Plus US$5 rebate on construction of second latrine pit Year 1- 5 (annual reward) US$0-10 each year based on following criteria • Toilet usage (verified) • Village toilet coverage (verified) • Completion of hygiene course • Presence of handwashing facility The plan would be supported by demand creation programs (CLTS, mass media), sanitation marketing programs to in- crease and improve the supply of low-cost sanitation goods and services, and microfinance programs to enable non- poor households to develop improved sanitation facilities. www.wsp.org 41 References ADB (2009) Proposed Asian Development Fund Grant – JMP (2008) Progress on drinking water and sanitation: spe- Kingdom of Cambodia: Second Rural Water Supply and Sani- cial focus on sanitation Geneva and New York: WHO-UNI- tation Sector Project: Report and recommendation of the Presi- CEF Joint Monitoring Programme for Water Supply and dent to the Board of Directors. Sanitation. Chapin J (2009) Design project: Sanitation marketing pilot Kar, K and Pasteur, K (2005) Subsidy or self-respect? Com- project, final report Phnom Penh: IDEO and IDE. munity-led total sanitation: an update on recent developments Brighton: University of Sussex, Institute of Development Cairncross S, and Valdmanis V (2006) Water Supply, Sani- Studies, IDS Working Paper 257. tation and Hygiene Promotion Chapter 41 pp. 771-792 in Jamison et al (2006) Disease control priorities in developing Knowledge Links (2007) Formative research: development of countries: Second edition Washington DC: The World Bank sanitation IEC manual for Himachal Pradesh Unpublished and the Oxford University Press. report. DHS (2006) Cambodia: Demographic and Health Survey Komives et al (2008) Water, electricity and the poor: who 2005 Maryland: ORC Macro and National Institute of Sta- benefits from utility subsidies? Washington DC: The World tistics. Bank. Fiszbein et al (2009) Conditional cash transfers: reducing pres- Mehta M (2008) Assessing microfinance for water and sanita- ent and future poverty Washington DC: The World Bank. tion: exploring opportunities for sustainable scaling up Bill & Melinda Gates Foundation, Final Report Government of India, Ministry of Rural Development, http://www.gatesfoundation.org/learning/Pages/microfi- Department of Drinking Water Supply Office Memoran- nance-for-water-and-sanitation.aspx. dum No. W-11037/6/2005-CRSP Revision of the unit cost of IHHLs under the Total Sanitation Campaign dated 21 Oc- Moraes et al (2003) Impact of drainage and sewerage on diar- tober 2008. rhoea in poor urban areas in Salvador, Brazil Transactions of the Royal Society of Tropical Medicine and Hygiene 97, Grosh et al (2008) For protection and promotion: the design pp.153-158. and implementation of effective safety nets Washington DC: The World Bank. MRD (2009) Community-Led Total Sanitation in Cambo- dia: Formative evaluation report Ministry of Rural Develop- Hutton et al (2009) Economics impacts of sanitation in Lao ment, Government of Cambodia, draft report. PDR: a five-country study conducted in Cambodia, Indonesia, Lao PDR, the Philippines and Vietnam under the Economics Prüss-Üstün and Corvalan (2006) in World Bank (2008) of Sanitation Initiative The World Bank, Water and Sanita- Environmental health and child survival: epidemiology, eco- tion Program. nomics, experiences Washington DC, The World Bank. JMP (2006b) Meeting the MDG drinking water and sanita- RMRCT (2007) Quantitative microbial risk based approach- tion target: the urban and rural challenge of the decade Ge- es to evaluate Nirmal Gram and Non Nirmal Gram Villages neva and New York: WHO-UNICEF Joint Monitoring of Rewa District, Madhya Pradesh UNICEF Project report. Programme for Water Supply and Sanitation. 42 www.wsp.org References Roberts M (2006) Demand assessment for sanitary latrines in rural and urban areas of Cambodia Phnom Penh: Interna- tional Development Enterprises. Robinson A (2005) Scaling up rural sanitation in South Asia New Delhi: The World Bank, Water and Sanitation Pro- gram South Asia. Robinson A (2009) Sustainability and equity aspects of total sanitation programmes: a study of recent WaterAid-supported programmes in Nigeria London: WaterAid, Report. TARU (2008) Impact assessment of Nirmal Gram Puraskar awarded panchayats: final report UNICEF. Tremolet S, Perez E and Kolsky P (2009) Financing on-site sanitation for the poor: a global six-country comparative re- view and analysis The World Bank, Water and Sanitation Program, draft report. UNICEF (forthcoming) Lao PDR: Child well-being and disparities – health, nutrition, water sanitation hygiene, edu- cation and protection. Waterkeyn J and Waterkeyn A (2005) Taking PHAST the extra mile through Community Health Clubs: the AHEAD methodology Africa AHEAD website WHO (2001) Macroeconomics and Health: Investing in health for economic development Geneva: World Health Or- ganization. www.wsp.org 43 ANNEX 1 MEETINGS HELD IN CAMBODIA Siem Reap: 05-07 October, 2009 1. Karin Schelzig-Bloom, Social Sector Specialist, ADB 2. Jan Willem Rosenboom, WSP Cambodia Country Team Leader 3. ADB TS-RWSSP Provincial Management Unit, Siem Reap Province 4. ADB TS-RWSSP Provincial Management Unit, Kampong Thom Province 5. Mao Saray, ADB TS-RWSSP Project Director Phnom Penh, 08-23 October 2009 1. Lyn McLenan, Lien Aid Cambodia Program Manager 2. Danielle Pedi, WTO Singapore 3. Sahari Ani, Lien Aid Cambodia CEO 4. Phyrum Kov, WSP Cambodia Water Supply and Sanitation Analyst 5. Jobien Monster, WSP Cambodia independent consultant 6. Syvibola Oun, Plan Cambodia WES adviser 7. Geoff Revel, UNC Program Manager 8. Cordell Jacks, IDE Cambodia 9. Wan Maung, ADB TS-RWSSP Consultant Team Leader 10. Hilda Winarta, UNICEF Cambodia WASH program officer 11. David Hill, USAID Small and Medium Enterprise Development Project 12. Dr. Chea Samnang, Director of Rural Health Care, Ministry of Rural Development, Royal Government of Cambodia 13. Harold Alderman, World Bank Social Protection Specialist 14. Timothy Johnston, World Bank Cambodia Health Specialist 15. Rebecca Carter, World Bank Aid Coordination Specialist 16. Michelle Pendrick, World Bank independent consultant Also attended one-day World Bank organized seminar on conditional cash transfers (19 October 2009, Phnom Penh). 44 www.wsp.org