Water and Sanitation Linking Sustainability with Demand ... SUSTAINABILITY WITH DEMAND, GENDER AND POVERTY participation in water users committees also contributes to sustainability

Bruce GrossChristine van WijkNilanjana Mukherjee

Water andSanitationProgram

An internationalpartnership to helpthe poor gainsustained access toimproved water supplyand sanitation services

A study in community-managed watersupply projects in 15 countries

Linking Sustainabilitywith Demand,Gender and Poverty

LINKING SUSTAINABILITY WITH DEMAND, GENDER AND POVERTY

ii

Contents

Foreword iii

Preface v

Acknowledgements vii

Introduction 1

Key Findings: Gender- and Poverty-Sensitivity Pays off in Sustainability 2

Implications: Making Demand-Responsive Services More Sensitive to 2Gender and Poverty Aspects

Why Gender, Poverty and Demand? 4

Participatory Assessments 6

Communities and Their Services 12

A Ranking Produces Two Clusters of Communities 12

A Mixed Picture Emerges of the 88 Communities 15

Significant Findings 17

Factors Associated with Sustained Services 17

The Kind of Participation That Leads to Sustained Services 20

Gender- and Poverty-Sensitive Agency Policies and Practices 23

Implications for Development Practice 26

Towards a Framework for More Equitable and Sustained Services 29


iii

he World Bank�s policy research team onengendering development found that

gender inequality is more pervasive in poorcountries. In those countries, the disparities aregreatest among poorer people. Though in noregion of the world have women achieved fullequality, in poorer regions they continue to havefewer rights, less access to resources, and morelimited voice. Many of the disparities betweenwomen and men�in rights, resources, andvoice�are also disparities between the poorand the better off members of society. Thoughwomen and girls bear the brunt, these disparitiesreduce economic growth and constraindevelopment generally. The study team, whosefindings were recently published by The WorldBank in Engendering Development - throughGender Equality in Rights, Resources, andVoice, recommends a broad, three-part strategyto promote gender equality that will give allmembers of a society a better chance to escapefrom poverty.

Those of us working in water supply andsanitation can promote a number of the specificmeasures recommended in the study at themicro level by designing and implementingprojects that better take into account gender

Foreword

and pover ty perspectives. These includemeasures such as service delivery to facilitateequal access for women and men; promotinggender equality in access to productiveresources and earnings capacity; reducingpersonal costs to women of their household rolesto free up time for more productive activities;and strengthening women�s voice andparticipation.

This report, Linking Sustainability withDemand, Gender and Pover ty, presents theresults of a study of community-managedwater supply ser vices in 88 communitiesaround the world. It adds to the evidence thatgiving more attention to the service demandsof women as well as men, of the poormembers of a community as well as thosewho are better of f, pays off directly in termsof sustainability of the services. Providinggreater voice and choice to all during theprocess of service establishment results ingreater equity later�a better distribution ofthe burdens and benefits among communitymembers, and improved water supplies thatmore community members use in a health-promoting manner. Greater access of womenand poor people to training and to


participation in water users� committees alsocontributes to sustainability.

The report points to some specific actionsthat agencies can take to target women andpoor community members more effectively.Among them are participatory tools that leadto greater inclusion of often-excluded groupsin making decisions throughout the serviceplanning and implementation process. Themethodology developed to conduct the study�Methodology for Participatory Assessments withCommunities, Institutions and Policy Makerspublished by the Water and SanitationProgram�is a tool that agencies can use toplan for and monitor sustainability.

These two publications provide practicalguidance to help water supply and sanitationsector practitioners and policy makers targetwomen and poorer community members moreef fectively. Doing so will lead to moresustainable and equitable development thatcontributes to broader societal goals, includinggender equality and a higher quality of life forthe poor.

Nemat ShafikVice President

Private Sector Development andInfrastructure

The World Bank

iv


v

hen we began the ParticipatoryLearning and Action (PLA) Initiative

more than three years ago, we and the othermembers of the PLA team had two objectives:The first was to add to the knowledge thatgreater attention to gender pays off in increasedsustainability. We proposed to do this byconducting a series of rapid assessments inprojects known to have some participatory anddemand-responsive features, to see if moregender-sensitive demand was linked to betteroutcomes. We reasoned that agency officialsresponsible for water supply and sanitationservices would pay more attention to gender, ifthey had evidence that doing so pays off inbetter sustained services. As we started to planthe study, we immediately added poverty�poorand better off�as a dimension, recognizingthat poverty equals gender in importance asan element of a community�s heterogeneity.

Our second objective was a five-yearcapacity-building phase to use the knowledgeto help change the way programs and projectsare designed and implemented. It was onlywhen we discussed practical follow-up to theassessments that potential funding agenciesfound they could justify the relatively highestimated costs of the assessments, thenenvisaged for 14 countries.

Preface

Achieving these two objectives has provedto be more challenging, has taken longer, andhas cost more than we originally envisaged.Nevertheless, with this publication, we arecompleting the first objective, of addingempirical evidence that gender and povertytargeting matter to sustainability. Although wehave not attempted to show a causalrelationship, the data from the 88 communitiesin 15 countries are clear and the associationsare significant � projects that used more gender-and pover ty-sensitive demand-responsiveapproaches had results that were bettersustained. These communities also used theirservices more effectively, i.e., the majority usedthe improved water supply in a health-promoting manner. The findings also show thatgiving choices to more community groups�thepoor, better off, women and men�and lettingthem influence (or control) the process of serviceestablishment empowers them to later manageand sustain their services more effectively.

These findings are consistent with demand-responsiveness. Giving users the kind of servicesthey want and are willing to pay for makes formore satisfied users, services that continue tomeet users� demands over time, and aretherefore better sustained. This may not bestartling to some, but evidence from around the


vi

globe can help convince skeptics�andstrengthen the resolve of supporters�that theeffort to reach out to all major groups in poorcommunities is worth making.

Our second objective of using the results hasbeen underway for some time, since wepublished the methodology to conduct the 88community assessments, now called theMethodology for Par ticipatory Assessments(MPA). The MPA is a methodology that canmake projects more demand-responsive, withmethods to identify the poor and better offmembers of communities, and then to give themmore equitable and greater voice and choiceduring the process of service establishment. TheMPA has other features that make it a usefultool for communities, project and agency staffto plan for and monitor sustainability, access,use, benefits, and user satisfaction in a povertyand gender-specific manner.

The challenge now is to move ahead anduse the results towards the goal of bettersustained and more effectively used watersupply and sanitation services. We envisage aset of interconnected tasks:

❖ To apply the MPA to large-scale programs,without endangering its interactive, learningqualities, while adjusting it to the specificconditions and needs of the individualsetting. Applications in planning andevaluation are already formulated orunderway in several projects, and there isinterest from others.

❖ To develop sufficient local capacity thatproject and task managers can readily callupon, without each time having to incur thetime and expense of undertaking extensivetraining, and to institutionalize this capacityinto existing country and inter-country sectorsupport organizations. The MPA is amethodology with an analytical frameworkthat drives towards sustainability and thatembeds gender and poverty at every stage.Even experienced hands in participatory

methods need some training to use itproperly and to its full potential.

❖ To continue to learn from experience andincorporate the lessons in the MPA in orderto target interventions more precisely andequitably, thereby improving theireffectiveness, efficiency and benefits. Twoactivities are necessary here. First, MPApractitioners need the opportunity to meetperiodically to exchange experiences andlessons among themselves and with peerreviewers who are specialists inparticipatory planning and evaluation. Inthis way experiences and lessons can be fedback into the methodology and its qualitypreserved when adjustments are made andapplications are scaled up. Second, a globalcore group needs to monitor the furtherdevelopment and modification of the MPA,to ensure quality and preserve the integrityof the MPA as a tool that monitorssustainability, embeds gender and poverty,and encourages learning at all levelsbecause it uses participatory (rather thanextractive) methods.

❖ The MPA�s analytical framework can beapplied in other sectors in order to plan andmonitor community-managed sustainableservices, provided that sector-specificindicators are developed.

From the outset the PLA has been a multi-faceted collaborative effort: first between theWSP and the IRC, as a global activity overseenby a global core group involving staff of WSP,IRC, and several other organizations, and thenamong the communities, projects, and agencies(government, NGO, and external supportagencies) that participated in the assessments.The collaboration has been fruitful andsuccessful, and we hope to broaden it as wemove forward.

December 2000 Bruce GrossChristine van Wijk

Nilanjana Mukherjee


vii

he Participatory Learning and Action (PLA)Initiative harnessed the talents and

energies of agencies and individuals aroundthe globe. It was a collaborative effort of theWater and Sanitation Program (WSP) and theIRC International Water and Sanitation Centre.It was planned, managed and carried out underthe supervision of a core team from the WSPand IRC.

Core team members took responsibility foroverseeing the 88 community assessments andvarious institutional assessments conducted:Rose Lidonde and Noma Musabayene inEastern and Southern Africa; Rekha Dayal withShalini Sinha in South Asia; Christine van Wijkwith Jennifer Francis in West Africa and MariaLucia Borba in Latin America; and NilanjanaMukherjee with Indrawati Josodipoero andKaren Jonesy Jacob in East Asia and the Pacific.These assessments could not have been donewithout the enthusiastic and supportivecooperation of women, men and watercommittee members in the communities, and inthe many project, government and non-governmental agencies that were involved.Nilanjana Mukherjee designed the basicsequence of participatory assessment tools withpoverty-targeting. Christine van Wijk took thelead in the development and pilot testing of themethodology and kept statistical analysis ontrack. Rekha Dayal served as the fieldcoordinator. Bruce Gross was responsible foroverall task management and for keeping thePLA Initiative funded.

Data analysis added a new set of people tothe core team. Among them, Uday Mehta andAnurag Rohatgi developed the initial plan andorganized and collated the data. Special thanksgo to A.J. (Viju) James, who joined the teamlate, and without whose tenacity and skills instatistical analysis over many months, the globalresults might still be elusive.

This report owes much to Suzanne Reiff, whotask-managed most of the data analysis anddrafting stage, and pulled together much of theoriginal material. Thanks also to the peerreviewers whose incisive comments sharpenedthe report�s messages: Nina Shatifan, SocialDevelopment and Gender Adviser, AusAID;Wendy Wakeman, Senior CommunityDevelopment Specialist, The World Bank; andMike (Harvey) Garn, now-retired economicadvisor to The World Bank�s Water andSanitation Division.

Managers at both the WSP and the IRCshared the PLA vision and supported theinitiative, and many other staff members of bothorganizations helped throughout, in ways toonumerous to mention. Lastly, and perhaps mostimportantly, thanks to the WSP�s donors, inparticular the governments of Canada, TheNetherlands, Norway, Sweden, and the Africaregion of the United Nations DevelopmentProgramme, without whose generosity and insome cases, active participation, this work couldnot even have been contemplated.

Acknowledgements


Introduction

n October 1997 a group of specialists fromnational, bilateral and international

organizations working in water supply andsanitation came together out of a commonconcern: how to better integrate gender-sensitive approaches into water sector workin order to promote more effectively sustainedand used ser vices. The meet ing wasorganized jointly by the Water and SanitationProgram (WSP) and the Gender Network(GENNET) of the Water Supply andSanitation Collaborative Council (WSSCC).The group agreed that much progress had

been made with regards to raising awarenessof gender issues in the WSS sector, butidentified two key areas that seemed tohamper further progress in the field:

❖ lack of empirical data regarding the extentto which gender-sensitive project ap-proaches actually have a significant impacton the sustainability of services createdthrough projects, and

❖ lack of a suitable methodology to help taskand project managers and staff deal withgender in projects.

Two earlier studies1 had looked intoparticipation and project effectiveness anddemand-responsiveness and sustainability.While the studies had identified key elementsof a demand-responsive approach, they hadnot inquired into dimensions of gender orpoverty (also a key equity concern). Meetingparticipants agreed that a study designedspecifically to investigate the links betweensustainability, demand, gender and poverty inprojects around the world could help fill theempirical data gap.

1Narayan, Deepa, The Contribution of People�s Participation, The World Bank, July 1995. Katz, Travis and Jennifer Sara, Making Rural Water SupplySustainable: Recommendations from a Global Study, UNDP-World Bank Water and Sanitation Program, 1997.

What could bring about sustained access of the poor to improvedservices?

1


To that end, the global Participatory Learningand Action (PLA) Initiative was launched. Thisreport presents the findings of the PLA studies,which have yielded valuable lessons andpractical strategies to assist funding andimplementing agencies to deliver more demand-responsive water services. It also providesresearchers and practitioners with insights intoan innovative methodology for participatoryresearch and community mobilization.

Key Findings: Gender- andPoverty-Sensitivity Pays off inSustainability

Key findings of PLA initiative shed light onthe linkages between gender, poverty, anddemand-responsiveness. In particular, the PLAinitiative found that:

❖ Better sustained services are more effectivelyused, by the majority of community membersin a health-promoting manner. Sustainedservices may not always be effectively used,for example, if the improved service reachesonly better off households. If they are noteffectively used, however, the communitycannot capture the full benefits of theimproved service, in particular, improvedhealth and quality of life.

❖ Better sustained services are significantlyassociated with a better gender- andpover ty-sensitivity in the demand-responsiveness of projects, user influenceand control over project implementation,sharing of burdens and benefits duringoperations, and user satisfaction.

❖ Services are better sustained when projectsoffer informed choices to both women andmen, poor and better off, thus empoweringthem to influence the process of serviceestablishment.

❖ The greater the sharing of control amongcommunity members�both women andmen, poor and better of f�in themanagement of the service, the more likelythe users will perceive the value of thebenefits of the service to equal (or exceed)the costs. The more satisfied they are, thebetter they sustain their services.

❖ Agency policies and objectives influenceresults on the ground, but agency staffingand operational practices are even morestrongly associated with better serviceoutcomes.

Implications: MakingDemand-Responsive ServicesMore Sensitive to Gender andPoverty Aspects

The goal of community-managed watersupply and sanitation is to develop communitiesthat can sustain and use their services effectively.The past decade has seen a lot of progress inunderstanding how better to reach this goal byoffering more demand-responsive services Thefindings point to a number of importantimplications for programming that can help toaccelerate progress in this direction.

❖ The right policies can leverage agencypractice in the direction of demand-responsive approaches that are moregender- and poverty-sensitive, i.e., that offermore options and information about them,to all major user groups. Projects sponsoredby agencies that mandate demand-responsive and gender- and pover ty-focused approaches are more likely toincorporate those approaches than projectsfrom agencies that do not.

❖ Agencies need the right staff, skill sets, andpractices to make their interventions more

2


demand-responsive and gender- andpoverty-sensitive. The availability of bothtechnical and social staff and their workingtogether in teams, and an agency�s capacityto plan and monitor using gender- andpoverty-disaggregated data are among thekey factors associated with better results onthe ground. A supportive management andstaff training are also important.

❖ The more that gender- and poverty-sensitivedemand-responsive approaches can beused from the beginning in interactions withcommunities, the more the community hasthe opportunity to influence the servicedelivery process, and the more it helps tobuild community ownership and capacity tomanage the services effectively.

❖ Promoting cooperation and mutualunderstanding at and between al lstakeholder levels, from the users up toagency management, can contribute tosustainability. Facilitated meetings wherethe stakeholders can hear each others�wishes and concerns and learn about how

their actions affect others, can be valuableto foster dialogue at every stage of theproject.

❖ Participatory social mapping and welfare/wealth classification are valuable tools forplanning, monitoring and evaluating thedegree and equity of access to services andother project inputs and outcomes, for women,the poor, and other marginalized communitygroups. Welfare/wealth classification andsocial mapping give a cross-section of womenand men in project communities a means toplan and monitor services in which facilities,training and functions of decision-making andcontrol are more equitably divided.

The Methodology for Par ticipatoryAssessments (MPA) provides tools to makedemand-responsive approaches a practicalreality. The MPA integrates gender and povertyand uses participatory tools and techniques atcommunity, project agency, and policy levels,thereby enabling community women and menand project and agency staff to themselves assessthe sustainability of the services.

3


Why Gender, Poverty and Demand?

better off men, better off women, poor women,and poor men�are likely to have differentwants and needs. They may also differ incapacities and forms in which they can supportwater and sanitation: in kind or in cash, in lumpsums or installments, in equal amounts oraccording to their different capacities. In ademand-responsive approach, in exchange fortheir contributions, users also have a voice andchoice in technology and levels of service,service provider, financing arrangements andmanagement systems, arrangements for sharingbenefits and burdens, and decisions on serviceadjustment and expansions. To elicit andrespond to the demands of various usergroups for satisfactory and more sustainableservices, DRA must be sensitive to dimensionsof both gender and poverty.

This is so, in the first place, because menand women have dif ferent roles andresponsibilities in society. They may attachdifferent values to services and the benefitsto be derived from them. Consequently, theirdemand for and use of services and their

ender and pover ty are essentialdimensions of demand-responsiveness

in all water and sanitation projects. A demand-responsive approach (DRA) takes into accountthat different communities and user groups�

4


economic behaviors differ. Similarly, the poorand the well-off have different expectationsfrom their services, different needs anddemands. In addition, they have verydifferent levels of influence on communitydecis ion-making processes. In mostdeveloping countries, the poor comprise themajority in communities, but they and womenlack the voice and power to make themselvesheard and heeded. For the same reasons,

5

they often do not benefit to the same extentfrom water projects as do men and betteroff households. Their burdens may increasedisproportionately with the improved service,for example, when they put more labor intomaintaining the system than their more affluentneighbors who are busy with income-generating activities. Their willingness to sustainthe service may be jeopardized if they perceivethe costs of the service to outweigh its benefits.


Participatory Assessments

o shed more light on the linkages betweensustained services and project approaches

that are more demand-responsive and gender-and poverty-sensitive, the WSP, in collaborationwith the IRC conducted participatoryassessments with 88 communities that hadmanaged and sustained their water supplysystems for three or more years.

A random sample of communities waschosen from 18 projects in 15 countries, fundedby a range of donors, NGOs, and governments.All the 88 communities included agreed toparticipate. The assessments were carried outfrom mid-1998 through mid-1999.

2Domestic use, as interpreted by project designers, usually includes the use of water for drinking, cooking, clothes-washing, cleaning and personalhygiene. Productive uses within the household are seldom considered explicitly.

The methodology developed to conduct theassessments assumes that the objective ofcommunity-managed water supply andsanitation projects is to create a service that isnot only effectively sustained, but is alsoeffectively used by the members of thecommunity. Effectively sustained service isdefined as a service that regularly and reliablyprovides enough water of an acceptable quality,as perceived by users as well as authorities, forat least domestic use2 . Breakdowns are rareand repairs are rapid (within 48 hours), andlocal financing covers at least the regular costsof operation, maintenance, and repairs.Effectively used service is defined as thecombination of the percentage of householdswith easy access to the improved water supply;the percentage always using the improved watersupply at least for drinking; and theenvironmentally sound use of the water system(presence of proper drainage and lack ofstagnant water).

A working hypothesis was developed, basedupon findings from previous studies, informalevidence, and years of experience from thefield: communities better sustain their services

Identifying the poor within each community and their access toservices is the first step in MPA.

6


when projects are more participatory, demand-responsive, and gender- and poverty-sensitive.Effectively sustained services and effectivelyused services thus are the study�s two dependentvariables. They were defined and assessedseparately so that relationships between themcould be tested, as well as between each ofthem and the study�s independent variables.

Participatory approaches involve peopleactively in making choices about their services.Gender- and poverty-sensitive approaches gofurther. A gender-sensitive approach givesequal access to project inputs and processesrelevant for the future ser vice, such asinformation, training, new functions and jobs,and decision-making to both women and men.In addition, a gender-sensitive approach paysattention to the distribution of work, resources

Targeting the Poor through Social Mapping andWelfare/wealth Classification

During the first day in a community, community men and womendraw a social map of the community. The map shows local roads,paths, community facilities, water points, homes and so on.Community members, with the help of the assessment team,

define welfare categories�better off, intermediate, worse off�based upon their ownperceptions and criteria for defining socio-economic status in their community, for example,type of housing, possessions, livelihood, etc. They then mark the location of the houses orcompounds of each category on the map. Clustering of each type of household on the maphelps identify neighborhoods that are predominantly better off, middle-level or poor. Themapping process in large communities is similar, but localities, not individual houses, aremapped by economic status.

The completed social map guides selection of focus groups, to ensure that both poorand better off households are included. It serves as a graphic guide to location of existingfacilities and to planning for improvements to ensure that poor households andneighborhoods are not overlooked, as the system is designed and sites chosen for locationof the public facilities. It can then be used to monitor physical implementation and theextent to which the improved services provide equitable access to poorer households withinthese communities. The combination of gender and poverty shows up by marking in whichhouseholds women or men have received training and hold functions or paid jobs.

and benefits between women and men, suchas inputs given during construction of theservice and use of water for productivepurposes within the household. To be gender-sensitive, projects must be prepared to dealwith women and men separately, if, forexample, women cannot speak up inmeetings with men.

A poverty-sensitive approach seeks to givea better voice and choice to the poor inrelationship with the better off members of acommunity and increase the balance betweenthe poor and the better off members of acommunity in decision-making. �Poverty�projects frequently target all communitymembers as �poor� if the community itself ispoor. Doing so misses real difference in welfare,and so in power and voice. Communities are

7


not, as Alice Welbourn so aptly observed,�homogeneous blobs.�3 They are tapestries ofdifferences in wealth, culture, religion, ethnicbackground�any or all of which have animpact on a group�s ability to be included indecisions that affect them. A poverty-sensitiveapproach identifies the poor and the better offthrough welfare classification and determineswhere they are physically located in thecommunity. It then seeks to ensure that, as theservice is planned, implemented and managed,the poor receive equal access to information,decision-making, conveniently located facilities,paid construction work, membership on thewater committee, and so on.

To assess the interrelationships betweengender and poverty perspectives in waterprojects, an analytical framework was developedthat looks backward from the present to the timeof service establishment. The analytical frameworkassumes that the degree to which a communitysustains an installed water and sanitation service(dependent variable A) and the degree to whichits population�male and female, poor and betteroff�effectively uses the service (dependentvariable B) are positively related to fiveindependent variable clusters:

C. The degree to which the service meets thedemands (both at service establishment andat time of assessment) of the majorpopulation categories�men and women,poor and better off,

D. The way in which burdens and benefits ofthe service are shared between men andwomen, poor and better off, and

E. The degree of gender- and poverty-sensitiveparticipation in the establishment andmanagement of the service.

Both the dependent variables A and B andindependent variables C, D, and E are assessedat the field level, in interactions with womenand men water users from the community, thecommunity water management organization,and project field staff. Gender and povertyperspective have been built into the analysis intwo ways: first, within the indicators andvariables; and second, by enabling groups ofpoor and better off women and men toparticipate, together as well as in separategroups, in the data collection and analysis atthe community level.

The analytical framework further examineswhat agency factors helped or hindered theprocess of service establishment, and whatpolicy factors in turn influenced the choice ofthe agency�s approaches. The institutional-levelindependent variable F is assessed atstakeholder meetings at district or province levelthat involve project staff and project managers,along with community members and socialintermediaries. The policy-level variable G isassessed at central level, in separate meetingswith policy makers, national level projectpersonnel, funding agency representativesand through document review. Theunderlying assumption is that the degree ofservice sustainability is positively associatedwith:

F. Institutional support for demand-responsiveand gender- and pover ty-sensitiveparticipation, and

G. The presence and application of demand-responsive and gender- and pover ty-sensitive policies in the project and the sector.

Table 1 shows the variable clusters and theindicators, which number 28 in total. The

3 PRA and Gender Workshop, Summary of plenary discussions, 6-7 December 1993. IIED/IDS, Sussex. mimeo. Gender and poverty are the two mostprevalent heterogeneity factors. Others such as religion or ethnicity may, in any given community, be just as important in determining whether a group hasa voice. Projects need to be alert to these factors and take steps to ensure excluded groups are properly included during the service planning andimplementation process.

8


Figure 1: Analytical Framework for the Assessments

analytical framework for the assessments, thevariables, and the indicators were initiallyconceived by the global PLA team, discussedand refined at a workshop with water-sanitationsector practitioners and specialists in genderand participatory approaches from Asia andAfrica, at Bangalore, India, and then pilot testedin Kerala, India.

This methodology with which the study wascarried out is now called the Methodology forParticipatory Assessments (MPA)4 . While itbuilds on earlier methods, notably thosepublished by WHO and Deepa Narayan of

UNDP/PROWESS5 , it is a new methodologyfor the participatory assessment of sustainability,use and gender- and poverty-sensitive anddemand-responsive participation for planning,monitoring or evaluation. The MPA usesparticipatory tools and techniques atcommunity, project agency, and policy levels. Itenables community women and men, andproject staf f themselves to assess thesustainability of the services with gender andpoverty aspects integrated into it. As such, itsvalue goes beyond its use in a single study. TheMPA also permits the data collected usingparticipatory techniques to be coded into

9

4 Rekha Dayal, Christine van Wijk, and Nilanjana Mukherjee, Methodology for Participatory Assessments with Communities, Institutions, and PolicyMakers, Water and Sanitation Program, March 2000. It contains detailed information that is not presented here, including a much fuller description of theassessment process.5World Health Organisation, 1983. Minimum Evaluation Procedures. Geneva, Switzerland, WHO. Narayan-Parker, Deepa. 1990. Participatory Evaluation:Tools for Managing Change in Water and Sanitation, New York: UNDP/PROWESS.


10

Note: �M/W� refers to women and men, and �R/P� to rich and poor.

A.Effectively Sustained

Services

B.Effectively Used

Services

C.Demand-responsive

Services

D.Gender and Poverty

Perspective in Sharingof Burdens and

Benefits

E.Participation in Service

Establishment andOperation

F.Institutional Support

for Gender- andPoverty-sensitive,

Demand-responsiveParticipation

G.Policy Support for

Gender- and Poverty-sensitive, Demand-

responsiveParticipation

A1 System Quality❖ Construction matches design; quality of materials and workmanshipA2 Effective Functioning❖ Service operation in terms of water quantity, quality, reliability, and predictability for

M/W, R/PA3 Effective Financing❖ Coverage of investment and/or recurrent costs❖ Universality and timeliness of payments by usersA4 Effective Management❖ Level and timeliness of repairs for M/W, R/P❖ Quality of budgeting and keeping accounts

B Hygienic and Environmentally sound Use by Majority❖ Proportion and nature of population using the service by M/W, R/P❖ Degree of improvement of water use habits❖ Presence and state of waste water disposal provisions for R/P

C1 Initial User Demand❖ Type and proportion of contribution by M/W, R/P at startC2 Project Responsiveness to Demand❖ User voice and choice in planning and by M/W, R/PC3 User Satisfaction❖ Satisfaction of user demands, of M/W, R/P❖ Ratio of user-perceived cost-benefits for M/W, R/P

D1 Gender and Poverty Focus at Start❖ Nature of community payments in the establishment of the service (degree of equity in the

system for payment, by M/W, R/P)❖ Cost sharing/contributions sharing between and within households at the time of service

establishment, by M/W, R/PD2 Gender and Poverty Focus during Operation❖ Division of skilled/unskilled and paid/unpaid labor between M/W, R/P in operation and

maintenance❖ Division of functions and decision-making between M/W, R/P

E1 Equity in Community Management❖ Responsibilities for maintenance, management, repairE2 Participation with Empowerment❖ Degree of community monitoring & control in construction schedules, quality of works,

and community contributions, by M/W❖ Composition, status, rules and tools of control of managing committee, : extent to which

present and gender & poverty-conscious❖ Types of skills created and practiced among M/W, R/P❖ Transparency in accounts (shared with M/W, R/P)

F1 Enabling Organizational System❖ Explicitness of Sustainability, Equity, Gender-sensitivity, Demand-responsiveness and

Community Management in project objectives, strategies and performance evaluation criteria❖ Gender- and class-disaggregated planning and monitoring systems in operation❖ Expertise as reflected in the type of agencies involved, field teams, and team approachF2 Enabling Organizational Climate❖ Managerial support, staff capacity building, and staff performance incentives

G Supportive Sector Policy and Strategy❖ National sector policy for water and sanitation present with Sustainability and Equity as

explicit goals❖ Degree to which national sector strategies are present to guide the achievement of policy

goals and incorporate community participation and management; gender sensitivity andgender equity; poverty sensitivity; and demand-responsiveness in sector projects

Table 1: MPA Indicators for Water Supply Services

Variables Indicators and Sub-indicators


Ordinal Data in the Assessments

Most community-level information generated in the assessments was ordinal. It differs frominterval or ratio data in that it can be ranked, but it should be noted that the differencesbetween scores are not proportional. For example, in Table 1 there are four indicatorsused to assess dependent variable A�effectively sustained services. One indicator ofeffective financing was an assessment of the revenue collected, scored as follows:

0 = no income or income covers only part of O&M costs

1 = income covers all O&M costs

2 = income covers all O&M costs and system expansion

3 = income covers all O&M costs, system expansion, andmajor replacement costs.

A ranking of 3 is most desirable for financial sustainability. While 3 is a higher rankingthan 1, it is not necessarily 3 times better than a ranking of 1.

ordinal scales, which can be fed into a programor project database and statistically analyzed.It thus provides a multi-level management toolfor monitoring, adjusting, and predictingsustainability with a gender and poverty focuson community-managed water services in largeinfrastructure projects and programs.

Substantial effor t and time went intochecking and rechecking the data, and anumber of discrepancies were found andcorrected several months after conclusion ofthe field work. The global sample of 88communities was sufficiently large for thepurpose of s tat is t ical analysis. Thecorrelations that appear in this report werecalculated using non-parametric statistics, theSpearman�s rho. Descriptive statistics suchas frequencies and cross-tabulations wereparticularly valuable in producing country

and regional reports when sample sizes weretoo small for correlations to be run.

The correlations found show associations,not causation, between variables. Thus, forexample, if better sustained services are foundto be significantly associated with projects thatwere more demand-responsive (as was thecase), it does not mean that demand-responsiveapproaches will necessarily lead to, or �cause,�sustainability. The study did not attempt to showcausation by showing that projects that werenot demand-responsive were not sustained. Thenumber and nature of the significantassociations found, however, add measurablyto the growing evidence supporting thehypothesis, namely, that projects that are moreparticipatory, demand-responsive, and gender-and poverty-sensitive do indeed lead to betteroutcomes.

11


component. In 54 of them the local sanitationhad been improved, mainly by installinghousehold latrines. One community declined tobe included, but did take part in water serviceanalysis 6 .

A Ranking Produces TwoClusters of Communities

A ranking of communities from top to bottomshows communities differed markedly in howthey scored on demand, gender, and povertyindicators. Ranking communities from high tolow, according to their score on sustained andeffectively used services (dependent variablesA and B), produced a cluster of 13 closelyscoring communities at the top and another 13at the bottom.

Top Cluster of 13 Communities. The top13 communities also scored higher onparticipation, gender- and poverty-sensitivity,and on services based on demand and responseto demand. They present a clear picture of themore successful services in the sample.

Communities and Their Services

ommunities that took part in theassessments came from 18 different

projects. They had many dif ferent watertechnologies that were clustered into threegeneral levels as shown in Table 2. Theser vices had been managed by thecommunities for between 3 and 25 years andfor an average of 5.5 years.

Eight of the 18 projects, with a total of 55communities, also included a sanitation

Why do some communities sustain their services better than otherones?

6The PLA assessments established indicators for sanitation and set out to test the same hypothesis as for water. The results showed that the sanitation servicesin the sample were quite different from the water services: with some exceptions, they had not been established in response to demand and were notmanaged by the communities. Rather, sanitation frequently was taken up as a condition of getting water. The exceptions were 10 sanitation services inKerala, India that were established in response to demand, and tended to be more gender- and poverty-sensitive than the water services in the samecommunities. Given the situation globally, however, there were insufficient numbers of communities to test the hypothesis, or the indicators proposed, andno further analysis has been done in this study for sanitation services.

12


Table 2: Participating Communities

*National or local governments, or both, provided some type of funding support to all projects.**Key:Project Technology: A = Simple (single point, gravity, or private standpost)

B = Middle (household connection and private standposts but no treatment)C = Complex (various connection types plus treatment)

Project Type 1 = Water supply only2 = Both water supply and sanitation

13

Country Project External Number of TechnologyFunding Agency* Communities Type**

LATIN AMERICA

Bolivia PROSABAR Regional Development 4 B1 (3)Corporation of Santa Cruz, C1 (1)Social Investment Fund,World Bank, JICA

Colombia TRANSCOL DGIS, Ministry of Health 4 C1 (4)Ecuador ETAPA CARE 4 C2 (4)Peru APRISABAC SDC, NEDA 4 A2 (2), C2 (2)

EAST AND

SOUTHERN AFRICAKenya Kwale WSS SIDA, UNDP/World Bank 4 A1 (2), B1 (2)

ProjectMalawi Malawi National DANIDA, CIDA, World Vision, 9 B1 (9)

Rural Water Save the ChildrenSupply Program

South Africa Tisane and Laaste African Development Bank; 2 B1 (1), C1 (1)Hoop Water Community-Based WS ProjectProject

Zambia Northern & Western NORAD, Irish Aid 4 A1 (2), B1 (2)Province WaterSupply Program

WEST AFRICA

Cameroon Pro Village II KfW 4 A1 (4)HELVETAS Swiss SATA HELVETAS, African 4 B1 (3) C1 (1)Association Rural Development Fund, British HighWater Supply CommissionProgram

Ghana COWAP CIDA, UNDP/World Bank 4 A2 (2), C2 (2)Volta Region DANIDA 4 A1 (4)Water Supply andSanitation Project

EAST ASIA

Indonesia WSSLIC World Bank 4 A2 (1), B2 (3)FLOWS AusAID 4 A2 (2), B2 (2)

Philippines Central Visayas AusAID 5 B2 (5)Water & SanitationProject

SOUTH ASIA

India Kerala CWSSP DGIS, DANIDA 4 B2 (10)Nepal RWSS Project ADB 10 A2 (5), B2 (5)Sri Lanka CWSSP World Bank, AusAID 10 A2 (7), B2 (3)

Total number of services 88


rated water quantity as sufficient in onlyseven communities. In five, the quantity wasinsufficient to meet women�s needs forhousehold and productive uses. Waterquality was rated as good in all but onecommunity.

❖ Regularity and predictability of the supplywas good for women in all but twocommunities, and year-round water supplywas available in all but three communities.

❖ Four communities covered at least theoperation and maintenance costs of theservice, and three made profit (data coverlast three years). Two could not meet evenfull operations and maintenance costs (datamissing in four cases).

❖ Ten communities made budgets (one did not,two missing data), and ten shared accountswith both women and men users.

Bottom cluster of 13 communities. Thebottom 13 communities with the lowest scoreson the combination of sustained services andeffective use also scored poorly on sustainedservices alone. They also fared poorly ongender and poverty and various managementindicators.

❖ According to the users, the service had beeninitiated by all in only three communities.The others had been initiated by the project(four cases), local leaders (four cases) orthe local male or female population (onecase each). In the top 13, over 90 percentwere initiated by the people and involvedwomen and men.

❖ In decisions taken during the planningprocess, local women and men had muchless influence on technology and serviceslevels in 23 percent of the cases, and inlocation of facilities and local maintenanceand financing systems in 31 percent of the

❖ In 12 cases the community itself had madethe decision to initiate the project, eitherduring a meeting of community women andmen (5 cases), or during a meeting ofcommunity leaders with women and men(7 cases).

❖ Both men and women felt they participatedin choices about technology or service leveland location of facilities (11 cases); localmanagement organization (all cases); andlocal maintenance (12 cases) and financingarrangements (12 cases).

❖ All groups�the better off, poor, men, andwomen�participated in planning decisionsfor the service.

❖ Women participated on all of thecommittees, but one. Six committees weregender-balanced (female members between40 and 75 percent). In four cases womenwere over-represented, and in two they werepresent but under-represented.

❖ Poor and better off women felt they couldinfluence decisions about servicemanagement in all but one community,which had, surprisingly, a gender-balancedservice management committee. In the casewith no women on the committee, womenstill felt they had a voice since majordecisions were made during all-userassemblies.

❖ In 60 percent communities, users felt thatservice and benefits were worth theircontributions. There were hardly anydifferences between poor and better off,women or men.

❖ Despite the generally high marks on specificaspects, only 25 percent of the communitiesin the top scores had a service that wasadequate on all points. The amount of wateravailable was especially problematic. Users

14


cases. In only half of the cases, women andmen participated in choosing the local watercommittee.

❖ On average, planning decisions were madedemocratically with a process that includedwomen�s participation in only 34 percentof the communities with the lowest results.This contrasted with an average of 88percent for communities with the best resultson the combination of the two of thevariables: sustained services and effectiveuse.

❖ Only seven communities had receivedtraining, compared to all in the top thirteencommunities.

❖ While 61 percent of the best performingservice communities had either a genderbalance or more women in the local servicemanagement organization, only 30 percentof the worst performing services had acomposition that could be called gender-balanced. None had more women thanmen. Indeed, in almost half, women wereunderrepresented (compared to 15 percentin the best performing services) and in threecommittees there were no women (there wasone without women in the top group). In onlytwo cases did poor and better off womenconsider that women participated equallywith men in decision-making.

❖ Not surprisingly, in the lowest scoringcommunities, users were generallydissatisfied with their service. In ten of the13 cases, users were either disappointed intheir expectations or believed that they werecontributing more than they got, or both. Inhalf of the cases, women and men ratedthat the service met only half to threequarters of their various demands7.

❖ For all but one service, funds were insufficientto cover even the day to day operating costs.In one situation, income was sufficient to payfor these costs, but could not meet largerrepairs. Only one service organizationprepared an annual budget and only twoorganizations accounted for the service andtheir management to those that were usingthe service and were sustaining (or areexpected to sustain) its operation.

A Mixed Picture Emerges ofthe 88 Communities

Findings from the sample at large aremixed with respect to current operation andprospects for continued operation of thesystem. The results clearly indicate that bothcommunities and project implementingagencies need to understand better the multi-dimensional components of sustainability�technical, financial, institutional, social, andenvironmental. They must also work togetherbetter to address all of them more adequatelyin planning, operating, and maintainingcommunity water supply systems.

Systems function, but the repair record isnot encouraging for half the systems. At the timeof the assessment, all services were functional.The increased emphasis worldwide on andtraining for community operation andmaintenance is also apparently having apositive effect on community capacity to handlemaintenance and repair. In one third of theser vices, the users reported that localmaintenance and management staff usuallysucceeded in restoring the delivery of waterwithin 48 hours of an interruption.

However, half of the services generally didnot get repaired within two days when therewas breakdown and in another 10 percent

7These demands varied greatly. They might have been practical, such as closer and safer drinking water. They might have also been strategic, such as morewater and more time saved and the possibility of using these benefits for improving one�s position, e.g., for domestic production such as livestock rearingand small enterprises, or for leisure and rest.

15


repairs took more than a week. Since womennormally cannot store water for more than 48hours, in 60 percent of the services theyprobably had to go to other and often unsafesources until service was restored. Only a fewlocal management organizations took measuresto promote safe water uses in such cases.

Delivery of water is inadequate in manysystems. Of the 88 participating services, 60percent delivered water throughout the year,but 28 percent had seasonal shortages, and10 percent never supplied enough water to meetthe primary needs of the users. Access to wateris also not adequate. All or vir tually allhouseholds were being served in only 40percent of the communities.

Financing poses a major stumbling blockfor sustainability over the longer term. None ofthe systems received any operating subsidy.Thus, to survive over the longer term, all systemsneeded to collect sufficient revenue to meet atleast current operating costs. Yet, nearly halfthe services in the sample failed to meet eventhis basic requirement, or even plan for it. Lessthan half of the local service managementbodies prepared budgets for the operation andmaintenance of the service, and only slightly more

than half kept proper accounts. All services butone had a functional local managementorganization, usually a water committee (in onecase the local chief ousted the committee). In onethird, there were no rules guiding how thisorganization manages the services and how theusers use it. These findings suggest that there ismuch more work to be done if services are to befinancially sustainable over the longer run.

Use of water services has improved to thepoint that it can provide a basis for healthbenefits. Behavior changes by those with goodaccess to water are promising from a publichealth perspective. Of those who have beenserved, more than 75 percent used only theimproved water system at least for drinkingwater. This proportion has been cited as theminimum threshold necessary to achieve areduction in diarrhoeal diseases.8 However,these users also invariably used the drinkingwater source for production, such as growingvegetables, raising livestock, making shea9

butter, bricks, snacks and drinks for vending(mainly women), processing palm oil andcoffee, and growing cash crops (men). Theseactivities raise potential conflicts betweendrinking and economic uses of water whensupply is interrupted or limited.

8Esrey, Steven A, Complementary Strategies for Decreasing Diarrhoea Morbidity and Mortality: Water and Sanitation. Paper presented at the PanAmerican Health Organization�s Celebrating 25 Years of ORT. Washington D.C., PAHO, 1994.9Shea butter is made by women from the fruits of the Bassia tree, which resemble olives. It is valued as an article of food and of medicine, and is animportant commodity in parts of West and Central Africa. It keeps for a year without salt and is also popular because it is whiter, more solid, and morepleasant to the taste than butter made from milk.

16


success in keeping services running, as shownby the correlations in Figure 2.

1. Better sustained services are moreeffectively used by the majority ofcommunity members in a health-promoting manner.

Services that had better scores on the variousaspects of sustainability (dependent variable A,see the specific indicators in Table 1) scoredhigher on effective use. Effective use (dependentvariable B) was defined in terms of use by themajority in a health-promoting manner: ahigher percentage of households with accessto the service, more year-round use for drinkingwater, and more environmentally sound upkeepof water points. The use of more water�alsoan important indicator of improved hygiene�was not included because of the difficulty inmeasuring it.

Among the variable clusters used for dataanalysis, effective use was significantly relatedonly to sustained services. This suggests thatthe extent to which a community uses its serviceeffectively is substantially related to the extentto which services are effectively sustained, butnot to any of the independent variables. Indeed,

Service users in Indonesia assess how far the project process wassensitive to gender and poverty issues.

Significant Findings

Factors Associated withSustained Services

Five factors emerged from the assessmentsas significantly associated with community

17


Note: C2, C3, D2, and E2 are subsets of the independent variables C, D, and E that use one of more indicators to test for significant association with thedependent variables A and B. See Table 1.

Figure 2. Correlations: Factors Associated with Sustained Service

services can be effectively used only if they arefirst sustained, but there are examples wherethe converse is true, such as, when improvedservices reach only the better off households ina community. But unless services are effectivelyused, a community cannot capture the fullbenefits of the improved service.

However, some individual indicators makingup the effective use cluster did show significantassociations with some individual indicators ofthe independent variables. These are furtherexplained below.

Access of community households to serviceswas higher when women participated inmonitoring project implementation, and whenpotential user contributions during constructionwere more equitable (within communities as wellas within households). Equity in communitycontributions was scored from low to high: nocontribution, voluntary contribution as peoplesee fit, flat rate for all, and all contribute, butthe amounts were adjusted to their capacities(for example, communities sometimes chose toexempt certain very poor households or allowed

them to pay less or pay in installments). Withinhouseholds, highest equity in contribution meantlabor and material contributions by both menand women, whereas a low score meant laborcontribution only by women. Interestingly, mendid not always recognize women�s contributionof food and drinks for construction workers asa contribution of women�s labor or materials,whereas women invariably did so.

❖ Final access to services achieved by acommunity�s households was significantlyhigher when: the project was initiated bythe user households, and not by externalagencies or local leaders; men and womenfrom intended user households participatedin deciding sites for facilities, the localfinancing system, and the composition of alocal organization to represent users indecision-making; and the local managementorganization had a higher percentage ofpoorer community members.

❖ Health-promoting use of the improvedservices was higher when: womenparticipated in decisions regarding location

18


technique of matrix voting, local women and menbrought out how the decisions on the water supplyhad been made. In over 85 percent of the servicesat least one decision was taken in a democraticand gender-sensitive manner, in that it involvednot only the formal leadership but also commonwomen and men. Joint decisions were mostcommon on service management. In three-quarters of the communities, women and menjointly made decisions on the membership of localwater management organizations; in two-thirds,on the location of the facilities and the localfinancing arrangements. Participatory planningwas least common in the choice of technologyand/or service level. Yet such participationhappened in almost half of the services, andin 41 percent of the cases men and womentook part.

❖ The better the gender and poverty focusduring operations, the better the services aresustained. In the sample, gender and povertyfocus included the perception of the divisionof payments and labor within households,sharing of labor within communities, andthe sharing of functions and decisions withinthe water management committee.Communities that had a more equitabledivision of burdens and benefits alonggender and poverty lines during operationsscored higher on sustainability.

of water points; both women and men wereable to choose representatives for theirservice management organization; and thepoor constituted a higher proportion of thelocal management organization.

2. Better sustained services aresignificantly associated with a bettergender- and poverty-sensitivity indemand-responsive projects, userinfluence and control over projectimplementation, sharing of burdensand benefits during operations, anduser satisfaction.

Four of the independent variable clusterswere significantly associated with a community�skeeping its services in operation.

❖ The more demand-responsive the project,the better the service is sustained. Demand-responsive projects of fer communitymembers a say in local design and planningdecisions. The projects in the sample werenot consistent in terms of offering options.Typically, they tended to offer more financialand management options but fewer choicesin terms of technology types or levels ofservice.

When local choices were possible, who hada say, and who made choices? Using the PRA

Divisions of Burdens and Benefits between Men and Women

Communities still have a way to go to achieve a better division of burdens and benefitsbetween women and men. Voluntary maintenance work is either done by women or shared.When maintenance is paid, men do it. Upkeep of hygiene and sound environmentalconditions around waterpoints is almost exclusively the unpaid work of women. It is thiscleaning work that makes women tend to spend more time on the preservation of the waterservice than men. Service management is much more a voluntary job, and when it is, in 40percent of the cases women and men do it together. However, one in ten local watermanagement organisations have no women members, and in one third women are underrepresented. On the positive side, when management is paid, e.g., for the administrativework involved, women have a better share than they do for paid maintenance.

19


Participatory stakeholder meetings spark awareness and trigger change

At a stakeholder meeting on Flores Island, Indonesia, the participantswere evaluating the efficacy of gender training given to projectfunctionaries for enhancing women�s participation in projectprocesses. Several women representatives from user communities inthe project were present at the meeting. As project functionaries and

the manager were discussing ways to make the training more effective, the women beganto whisper agitatedly among themselves. The meeting facilitators encouraged them to discusswhat they were thinking among themselves. A while later, they gathered enough confidenceto address the whole meeting to point out that the problem of low participation of womenin project processes had much to do with the fact that project functionaries worked onlythrough the community leaders. Women were never directly invited or consulted by projectpersonnel. Community leaders traditionally ignored and excluded women from all decision-making. They said:

"If you are giving any gender-awareness training, give it first to our communityleaders! They are the ones that keep us out of everything. And why doesn�t theproject worker talk to us to find out what we want? Why must we always speakthrough our leader?"

The project manager and functionaries admitted during the meeting that they neededto change some things about the way they work. The women suggested what and how theycould change.

PLA Field Team, Flores

❖ The greater the participation withempowerment, the better the services aresustained. Services were better sustainedwhen communities par ticipated inestablishing the services not just bycontributing, but also by exercising a degreeof influence and control on the quality andpace of project implementation. This wasparticularly true where women participatedwith men in monitoring and control offinancing and construction. Local men andwomen have a wealth of local knowledgeabout what works and can be sustained andhave the greatest personal interest in jobswell done. This knowledge and interest canbest be utilized by projects when the projectagency is willing to consult and listen tothem, and give them some responsibility forand influence on the quality of projectimplementation.

❖ The more satisfied the users, the better theservices are sustained. The more that the usersperceived that the service had met theirdemands about its direct and indirect benefits,the greater the sustainability of the services.This included assessment of the degree towhich different user groups felt the benefitsfrom the service were worth what they paidfor them, in cash, kind, time, or labor. Thisfinding was not surprising, given the WSP�searlier global study of demand-responsiveness and sustainability, which alsofound that more satisfied users were morelikely to keep their services running.

The Kind of Participation ThatLeads to Sustained Services

In the past, water projects that aimed for�community participation� took a fairly narrow

20


21

view of participation. The project generallydecided about technology and service levels,sometimes decided sites and numbers offacilities in consultation with communities,proposed certain management structures anduser payments, and then got communityleadership to create the necessary structures andraise the necessary funds. Constructionproceeded, using contractors not accountableto communities. Constructed facilities werehanded over to user committees for operationand maintenance. Some training was given,usually to selected users or village functionaries,generally with no attention to gender or poverty.

The PLA assessments found that this passiveuser involvement is not the kind of participationthat leads to sustainability. Participation that linkswith sustainability provides an avenue for usersto influence, if not control, the process of estab-lishing services, and it does so not just for thelocal leaders, but for both men and women fromall major potential user groups. The significantfindings associated with the major variable clus-ters�gender and poverty in operations andparticipation with empowerment�are showngraphically in Figure 3.

1. Services are better sustained whenprojects offer informed choices to bothwomen and men, both poor and betteroff, thus empowering them to influencethe process of service establishment.

The impetus to move from a �supply� to�demand� orientation in the water sector beganin the 1990s following an internationalconference in Dublin and the Earth Summit inRio de Janeiro in 1992.10 The first generationof projects seeking to make the shift started byemphasizing willingness-to-pay�what acommunity is prepared to contribute towards

10From these two conferences emerged a set of principles for development in the water sector generally. Among the most important for water supply andsanitation are the recognition that water is both an economic and a social good and that water should be managed at the lowest appropriate level, withstakeholder�in particular women�s�involvement.11 Brown, Gillian and Richard Pollard, Experiences with implementing demand-responsive approaches: the case of Indonesia, Paper at Global conferenceon Community Water Supply and Sanitation, 1998, The World Bank, Washington D.C.

costs. Findings from the WSP�s rural watersupply study in Indonesia showed thatcommunity contributions were often mandatoryand decided by project personnel inconsultation with the village chief, not inconsultation with the users themselves. As aresult, people saw their contributions as a kindof tax, rather than a payment that they electedto make in return for a service over which theyhad some control.11 In such situations, a highinitial contribution could indicate a greater voiceof the better off in establishing services, withpossible inequities in later access by the poor.This could be an explanation for the onlysignificant negative correlation found in the PLAassessments between user demand (initialcontributions) and gender and poverty focusduring operation of the systems (see Figure 3).

A demand-responsive service needs to bedefined not only in terms of who in thecommunity and the households pay and howequitable the payment system is, but also interms of who makes informed choices inplanning services. These choices do not justconcern the type of system, but cover all keydecisions relevant for sustainability and use,such as how the system will be financed andmanaged and involves not only costs, but alsoaspects of health and social justice. Demand-responsiveness does not stop when services areimplemented, however. Only when theperceived benefits equal�or exceed�the coststo the users will the users go on supporting incash, kind and labor a water supply servicewhich they helped plan and install. The PLAassessments found sustained services to bestrongly associated with demand-responsiveness of projects (user voice andchoice in planning decisions), and usersatisfaction. No relationship was found,


however, between sustained services and initialuser demand as measured by initial usercontributions. This suggests that thesecontributions could have been viewed as a kindof tax rather than as contributions adjusted tooptions and benefits.

Users derive their sense of ownership andresponsibility for sustaining their services fromexercising control over planning, financing andconstructing the facilities, and then having theservices managed to their satisfaction. This kindof par ticipation, defined in this study asparticipation with empowerment, was found tobe clearly associated with sustained services. Thepattern of associations found suggests that genderand poverty focus at project start and the demand-responsiveness of the projects can bring aboutsuch participation with empowerment.

Despite good intentions, however, the extentto which the community is ultimately able to sustain

22

Figure 3. Factors Associated with Gender and Poverty Focus and withParticipation with Empowerment

Women and men assess how equitably burdens and benefits areshared once the new services are in place.

Note: C1, C2, D1, D2, E1 and E2 are subsets of the independent variable C, D, and E that use one or more indicators to test for significant associations withother variables. See Table 1.


its facilities also varies with two factors: how wellproject staff could and do inform them on theimplications of the various options, and to whatextent the project provides training both to menand women for technical, financial, management,and hygiene aspects of their services.

The findings of the study confirmed thisbroader relationship between contributions,influence over planning, and subsequentsatisfaction with the services and other perceivedeffects of the services. The more projects offerusers choice in technical, financial andmanagerial aspects and enable both women andmen to make decisions, the more they will laterexert control over service management.

2. Services are better sustained whenboth women and men, poor and betteroff, participate in the management ofthe service.

All but one of the 88 communities in theassessment had a water managementorganization. The presence of this organization,who participates in it, and the way in which itoperates are indicators of equity inmanagement of the service.

The variable cluster participation withempowerment includes indicators such as whetherthe management organization has a recognizedstatus and operates by rules; whether women andmen participate in monitoring and control of theconstruction process; the influence male andfemale members exert over the timing and qualityof system design and construction; whether ornot women and men in the committees andcommunity have been trained for technical,managerial and hygiene tasks; and whetherfinancial management is accounted for, and ifso, whether accounts are shared and with whom(local leaders only or with community women andmen).

This kind of community and user influence andcontrol over the services throughout the project

23

cycle turned out to be a central concept. Not onlywas it significantly related to sustained services,it was linked to initial demand, demand-responsiveness of project, user satisfaction, andequity in community-management in terms of whodoes the maintenance and management workand who gets paid for it. In other words, thegreater measure of control both women and menand better off and poor have in the managementof the service, the more likely the users willperceive the value of the benefits of the service toequal (or exceed) the costs, the more satisfiedthey are, and the better they sustain their services.

Gender- and Poverty-Sensitive Agency Policies andPractices

Enabling institutional environments andsupportive agency policies and practices arewidely accepted as essential ingredients insuccessful development programs. The PLAassessments looked explicitly at a number ofkey factors to better understand the linkagesbetween agency policies and practices andresults in the field. These factors were assessedin stakeholder meetings that brought togetherwomen and men who represented thecommunities in which services were establishedwith social and technical agency staff thatsupported the service establishment.

1. Agency policies and objectivesinfluence results on the ground.

Although policies must be translated intoactions through agency practices, foursignificant associations with results on theground emerged:

❖ The presence of agency objectives fordemand-responsive projects was significantlyrelated to the demand-responsiveness of theproject, to the empowerment of communitiesto influence and control their services, and touser satisfaction.


❖ Gender- and pover ty-focused agencypolicies were clearly associated with the levelof community participation at project start,community participation with empowermentduring operations, and higher usersatisfaction.

❖ Agency objectives for sustainability andgreater equity in access to service wereassociated with higher user satisfaction andgreater gender- and poverty-sensitivityduring operations.

❖ An agency policy encouraging communityownership and management producedgreater user satisfaction.

Figure 4 shows correlations between factorsassociated with gender and poverty and agencypolicies and practices.

2. Agency staffing patterns, skills, andteamwork are more significantlyassociated with service outcomes thanagency policies.

As might be expected, the ways in whichan agency operates and is staffed are moredirectly associated with outcomes at thecommunity level. Figure 5 shows the factors thathelp to explain the results on the ground. Amongthe most important factors include:

❖ Agency approaches for staff use in fieldoperations make a difference. When bothtechnical and social expertise was availableamong project staff and their availabilitywas combined with their use in a teamapproach, significant positive associationswere found with five other variable clusters:gender and poverty focus at start, initial

Figure 4. Correlations: Agency Policies and Results on the Ground

Note: F111, F112, F113, and F114 are subsets of the independent variable F1 used to test for significant associations with other variables. See Table 1.

24


user demand, demand-responsiveness ofprojects, user par ticipation withempowerment, and user satisfaction withservices. When only technical and socialstaff expertise was available but not usedin a team approach, the availability wasweakly correlated only with usersatisfaction.

❖ Supportive management and progressive12

agency staf f training programs areimportant ingredients linked to better resultson the ground. Progressive agencymanagement was the only aspect directlyand significantly (though weakly) associatedwith sustained services. It was betterassociated with the demand-responsivenessof projects, with gender- and pover ty-sensitivity at the outset of serviceestablishment, and with user satisfaction. Asmight be expected, progressive agency

management had a significant associationwith the presence of gender- and poverty-sensitive agency staff training. This in turnwas strongly associated with the presenceof incentives for staff to employ gender- andpoverty-sensitive approaches in their work.

❖ A planning system that enables an agencyto col lect gender- and poverty-disaggregated data is signif icantlyassociated with a gender and povertyfocus at the start and during operationsof the project. An agency�s capacity touse gender- and pover ty-sensi t iveapproaches depends, among otherthings, on a system that enables it tocollect and analyze gender- and poverty-disaggregated data. Without such asystem, it can neither effectively plan fornor monitor progress towards gender andpoverty objectives.

25

Figure 5. Correlations: Agency Skills and Practices and Results on theGround

Note: F121, F123, F124, F211, F212, and F213 are subsets of the independent variables F1 and F2 used to test for significant associations with othervariables. See Table 1.

12�Progressive� in this context means that training on social aspects, including on gender in the highest scoring option, is a regular feature of the programand involves the use of modern training methods, such as participatory tools.


both women and men and the poor and betteroff. Findings from the community assessmentsadd to the mounting evidence that theseapproaches result in services that have a greaterlikelihood of sustainability and lead to greaterequity if complemented by other inputs. Thefindings point to some important implicationsfor development practice:

1. The right policies can leverage agencypractices in the direction of demand-responsive approaches that are moregender- and poverty-sensitive, i.e., that offermore options and information to all majoruser groups. Although the linkages betweenpolicies and field results were less extensivethan those between agency practices andskills and field results, projects sponsoredby agencies that mandated demand-responsive and gender- and pover ty-focused approaches were more likely toincorporate those approaches than projectsfrom agencies that had not.

2. Agencies need the right staff, skill sets, andpractices to make their interventions moredemand-responsive and gender- andpoverty-sensitive. The emphasis on shiftingfrom supply-driven to demand-responsiveapproaches is less than a decade old. Whilesome water supply agencies now have

Implications for Development Practice

he goal of community-managed watersupply and sanitation is communities that

can sustain and will use their services effectively.The past decade has seen a lot of progress inunderstanding how better to reach this goal byoffering more demand-responsive services. Ademand-responsive service is one in which theusers, in exchange for making contributions (incash or kind), have a voice and choice intechnology, level of service, service provider,financing arrangements and managementsystems, arrangements for sharing benefits andburdens, and decisions on service adjustmentsand expansions.

Demand-responsive projects that are moregender- and poverty-sensitive give voice andchoice to more potential user groups including

Sector agencies need skills and approaches that offer choices tocommunities and let users' voices be heard.

26


experience with a first generation demand-responsive project, most are only nowbeginning to grapple with the implications.While some policies are in place, strategies,practices, staffing patterns, and trainingprograms still are geared to an earlier era.The associations in Figure 5 demonstrategraphically the importance of agency-levelcapacity and actions in shifting gears, fromsupply to demand:

❖ Achieving sustainability takes both socialand engineering staff, working together ina team approach that uses balancedinterventions at the project and communitylevels. Staff training is essential.

❖ Planning and monitoring systems should bein place that enable staff to employ demand-responsive, gender- and poverty-sensitiveapproaches at the project and communitylevels.

❖ Agencies, projects, and community-levelintermediaries need tools and methods, ofwhich the MPA is an example, and thecapacity to use them to undertake genderand poverty-disaggregated planning, datacollection, and analysis. This is necessary,at minimum, at the community and projectlevels. This includes assessing andresponding to the demands of oftenexcluded groups who are potentialconsumers such as the poor, women, andreligious and cultural minorities that oftenlive in separate enclaves. It also includes themeans to identify members of these groups,to provide them with information, and toengage them actively in order to give themboth voice and choice when serviceimprovements are contemplated.

3. The more that gender- and poverty-sensitiveand demand-responsive approaches can beused from the beginning in interactions withcommunities, the more the communitymembers have the opportunity to influencethe service delivery process, and the more

it helps to build community ownership andthe capacity to manage the serviceseffectively. This suggests that projects andprograms need to focus on empoweringcommunities to sustain their access in theprocesses of design and technology choiceand in the training of community members.Expecting that agencies can continue tocome back to served communities to replaceor expand services is unrealistic, inefficient,and reduces the resources available for theunserved poor.

A community�s capacity to participate withempowerment in the management of andcontrol over their service emerged as a centralconcept in the findings, and significantly relatedto demand-responsiveness, gender and povertyfocus at project start and later on, to usersatisfaction, and sustained services. Projectsthat were initially more gender- and poverty-sensitive in inviting and in meeting demandsfrom more segments of the community,performed better over time, and thecommunities were more empowered tolater take control of the management oftheir services.

Project agencies can contribute toempowerment throughout the project cycle bybringing out and being responsive to differentdemands and by using gender- and poverty-sensitive interventions at every stage. Both menand women in communities need to understandthe implications of the potential technologiesand service levels�in terms of cost, ease ofmaintenance, repayment and management,extent of access (and what this implies in socialand health respect), their differing potential forupgrading, their implications for the uses ofwater for other than domestic purposes. Theyneed to make informed choices on the locallybest-fitting forms of water management andfinancing, assisted to set up rules for itsoperation and for managing the accounts. Earlyestablishment and training of the water userscommittee also helps to build communityownership and control from the beginning.

27


intermediaries or project staff. If findingsfrom participatory assessments areavailable, they will reinforce the communityvoices and give the communities an evenstronger opportunity to make themselvesheard.

6. Participatory social mapping and welfare/wealth classification are valuable tools forplanning, monitoring and evaluating thedegree and equity of access to services andother project inputs and outcomes, forwomen, the poor, and other marginalizedcommunity groups. Communities are nothomogeneous, and projects must havemeans to identify and engage a cross-sectionof community members throughout theprocess of service establishment. The moreparticipatory the social mapping process,the more awareness is created of theimportance of finding equitable ways ofproviding access to women, the poor, andother marginalized groups.

The Methodology for Par ticipatoryAssessments (MPA) developed to conduct theassessments has demonstrated that it cangenerate information useful to all stakeholdersfor enhancing sustainability. It also providestools to make demand-responsive approachesa practical reality. The MPA integrates genderand poverty and uses participatory tools andtechniques at community, project agency, andpolicy levels, thereby enabling communitywomen and men and project and agency staffthemselves to assess the sustainability of theservices. It also permits the data collectedusing participatory techniques to be codedinto ordinal scales, which can be used indatabases and statistically analyzed. Appliedsystematical ly and in sequence, theparticipatory tools and techniques of the MPAare a way for communities and agencies toplan for and monitor sustainability, access,use, benefits and user satisfaction in apoverty and gender-specific manner.

This kind of participation leads tosustainability. It is also the kind of participationthat might require more time and more flexibleimplementation schedules, so budget provisions,timelines, training programs, and selection ofconsultants and NGOs have to be plannedaccordingly.

4. Projects should strive to provide moreoptions to all major user groups in thecommunity. Doing so will make for moresatisfied users, lead to a better division ofthe burdens and benefits of the water andsanitation service, and promote greatersustainability. Community groups often havedifferent demands for service. Users alsosupport water services for more than wateralone, and women and men have differentreasons for doing so. The ways in whichthey do so, and the resources they have alsovary, for poor and better off women, andfor poor and better off men. These demandshave to be considered and evaluated forthe various user groups. The project shouldrespond to these different patterns ofdemands by giving various groups choicesrelated to technology, level of service, andthe purposes for which water is used. Wherechoices cannot be accommodated, forexample, because the desired solution is nottechnically feasible, the users need to knowthe reasons, so they can make anotherchoice, or opt out of the project altogether.

5. Cooperation and mutual understanding atall stakeholder levels�from the users up toagency management�contribute tosustainability. Facilitated meetings duringwhich stakeholders from different groupscan visualize their needs and concerns andunderstand how their demands interrelatewith others can help to foster dialogue atevery stage of the project. In particular, theycan give communities a much more directsay in the process than they would have ifthey are only represented from afar by

28


Towards a Framework for MoreEquitable and Sustained Services

he findings from the communities suggeststrongly that demand-responsive

approaches that integrate gender and povertyare the route to sustainability of community-managed water supply and sanitation services.From the findings and the significantassociations among the dependent andindependent variables, the following frameworkfor sustained and more equitable community-managed water supply services emerges.

Figure 6: A Framework for Sustained and More Equitable Services

Services are better sustained and have a greater health impactwhen they are equitably accessed by all sections of the community.

29

linked significantly to effectively sustainedservices, and to equity through gender andpoverty approaches in eliciting demand and inmanaging the service. Agency staff, skills, andpractices have a significant impact on serviceoutcomes. Agency policies guide businessprocesses and are an indispensable underlyingsupport, one that will help attract the rightmanagement and help obtain the resourcesnecessary to bring about changes in agencyoperation.

The challenge at hand is now how totranslate this framework into practice. The nextsteps will be to apply the MPA to large-scaleprograms with the ultimate goal of leading tobetter sustained and more effectively used watersupply and sanitation services.

The framework is related to the analyticalframework developed for the MPA, but addsimproved quality of life of the poor as anultimate goal. Although the PLA study did notassess the contribution of water and sanitationservices to quality of life, water and sanitationare basic services. Those who have them mustbe seen to enjoy a better life than those without.

Sustained services are more likely to resultfrom project interventions when they respondto the demands of all potential users�the poor,better off, women, and men�and empower theusers to take greater control over their servicesthroughout the cycle, from design to operationand management.

Effective use of the services, which isnecessary for improved community health, is

30


Water and Sanitation ProgramThe World BankWater Supply and Sanitation Division1818 H Street, NWWashington, DC 20433

Telephone: (202) 473-9785Fax: (202) 522-3313/3228E-mail: [email protected]: http://www.wsp.org

The World Bank does not accept responsibility for the views expressed herein, whichare those of the authors and should not be attributed to The World Bank or its affiliatedorganizations. The findings, interpretation and conclusions are the result of a researchsupported by the Bank. The designations employed and the presentation of thematerial are solely for the convenience of the reader and do not imply the expressionof any legal opinion whatsoever on the part of The World Bank or its affiliatesconcerning the legal status of any country, territory, city, area, or its authorities, orconcerning the delimitations of its boundaries or national affiliations.

January 2001

The Water and Sanitation Program is aninternational partnership to help thepoor gain sustained access to improvedwater supply and sanitation services. TheProgram�s main funding partners are thegovernments of Australia, Belgium,Canada, Denmark, Germany, Italy,Japan, Luxembourg, the Netherlands,Norway, Sweden, Switzerland, and theUnited Kingdom; the United NationsDevelopment Programme and The WorldBank.

Designed and printed byRoots Advertising Services Pvt. Ltd.E-mail: [email protected]

IRC International Water and Sanitation CentrePO Box 2869, 2601 CW DelftThe Netherlands

Telephone: +31 (0) 15 291 2939Fax: +31 (0) 15 219 0955E-mail: [email protected]: http://www.irc.nl

Documents

Water and Sanitation Linking Sustainability with Demand ... SUSTAINABILITY WITH DEMAND, GENDER AND POVERTY participation in water users committees also contributes to sustainability