drinking water, sanitation, 19 wdr ch09.qxd 8/14/03 10:01 am …. drinking... · 2016-11-22 ·...

159

Drinking water, sanitation, sewage disposal,electricity, rural roads, and urban transportinfluence human development outcomes(crate 1.1). As with education and healthservices, the impact of infrastructure ser-vices on human development is direct (e.g.,reducing water-related diseases, which rankamong the top killers of children). Theimpact is also indirect, through economicgrowth.452 But like education and health,these services are failing poor people.

Focusing on water, sanitation, and elec-tricity services, this chapter uses the Report’sservice delivery framework to find out whyand to show how things might be improved.The reform lessons from these services, rep-resenting both network and non-networkservices, are also likely to apply to otherinfrastructure services.

For networked services, such as urbanwater and electricity, regulating providersand ensuring that poor people have access toaffordable services are the main reasons forgovernment intervention. This brings thelong route of accountability into play. Butpoor citizens have a weak voice because waterand electricity are particularly vulnerable topatronage politics. Providers end up beingmore accountable to policymakers than toclients, which breaks the long route ofaccountability.

The solution is to separate the policy-makers from the providers—and to makeproviders more responsive to clients. Dis-persing ownership through decentraliza-tion and private participation, promotingcompetition through benchmarking, ensur-ing alternative access by using independentproviders, and charging for services areways of separating policymakers fromproviders and strengthening compacts,client power, and voice.

In rural network and non-network set-tings, community and self-provision domi-nate. The policymaker as standard setterand capacity builder in support of the clientis missing. To avoid ensuing problems, suchas arsenic in Bangladesh’s rural drinkingwater, policymakers need to support clientsin ensuring service quality and access.

Externalities in sanitation in rural, non-network settings are best contained within thevillage or community. So supply-side supportat the household level should be comple-mented with interventions at the communitylevel—be it information about hygiene orsubsidization of latrines—that are designed tospur household demand and create commu-nity peer pressure for behavior that internal-izes the externalities. In urban settings, wheredemand for sanitation services may be greater,property rights and facilitating privateresponse can support collective efforts.

The state of water and sanitation servicesAbout 2 of every 10 people in the developingworld were without access to safe water in2000; 5 of 10 lived without adequate sanita-tion; and 9 of 10 lived without their waste-water treated in any way.453 There have beengains, but despite the many global commit-ments, notably the U.N. Decade for Waterand Sanitation, access to water and sanitationlags far behind the milestones set in the1980s. Nor do aggregate trends in the 1990sgive comfort (figure 9.1). The share of peoplewith access to these services in Africa andAsia—where the world’s poor are concen-trated—has fallen, remained constant, orincreased only slowly.

Innumerable city and town studies con-firm the UN-Habitat Report’s key message

Drinking water, sanitation,and electricity

9c h a p t e r

19_WDR_Ch09.qxd 8/14/03 10:01 AM 59

that water and sanitation services are toooften failing communities.454 Full-pressure,“24-7” water supply remains a pipe dreamin many cities. Because a quarter to half(and more) of urban water supply remainsunaccounted for, many cities are turnedinto leaking buckets (figure 9.2). The lim-ited number of network access points mustbe widely shared, which dramaticallyincreases waiting times and often simplyoverwhelms the system. Rural infrastruc-ture often goes to seed: more than a third ofexisting rural infrastructure in South Asia isestimated to be dysfunctional.455

Poor people bear a disproportionateshare of the impact of inefficient water andsanitation services. Fewer poor people areconnected to a network. When they do haveaccess, the installation has to be sharedamong many more people (figure 9.3). Andthe prices they pay are among the highest,generally more than those paid by moreaffluent households connected to the pipedsystem (figure 9.4). The price differential ispartly a result of inefficiencies—theinequitable practice of subsidizing pipedwater, lack of scale economies for indepen-dent providers, or worse, providers takingadvantage of poor people’s lack of choice.But some of the price differential can alsoreflect the flexibility and convenience ofservices offered by independent pro-viders—no connection charges or access toquantities of water that are more affordablefor poor people.

Infrastructure and theaccountability framework for service deliveryCountries are trying different approaches toaddress failing water, sanitation, and electric-ity services. These include decentralizing tolocal governments, private sector participa-tion, regulatory reform, community-drivendevelopment, and small independentproviders. Some approaches try to make ser-vices work for poor people through targetedinterventions. Others seek to improve ser-vices overall—on the premise that makingservices work for all is necessary for makingthem work for poor people. The sameapproach has worked in one setting andfailed in another, and different approacheshave worked in seemingly the same setting.What is needed is a way to think about theinstitutional and political characteristics ofinfrastructure services to understand whatworks where and why.

Accountability in infrastructureservicesChapters 3–6 of this Report develop a frame-work for analyzing how well the actors in ser-vice delivery—clients and citizens, politiciansand policymakers, and service providers—

160 WORLD DEVELOPMENT REPORT 2004

Figure 9.1 Little progress in access to improved water and sanitation, 1990 and 2000

01990 2000

Africa

Urban

Rural

Asia1990 2000

Latin Americaand the

Caribbean

25

50

75

100

1990 20000

1990 2000Africa

Urban

Rural

Asia1990 2000

Latin Americaand the

Caribbean

25

50

75

100

1990 2000

Percent PercentWater supply coverage by region Sanitation coverage by region

Source: WHO, UNICEF, and Water Supply and Sanitation Collaborative Council (2000).

24 16 8

Dhaka

0 20 40 60

Phnom Penh

Beijing

Hong Kong

Jakarta

Manila

Bangkok

Calcutta

Chennai

Delhi

Kathmandu

Karachi

Colombo

Nairobi

Nakuru

Hours of wateravailable in a day

Percent of waterunaccounted for

Source: Human Settlements Program (2003).

Figure 9.2 24-hour water: a pipe dream

19_WDR_Ch09.qxd 8/14/03 10:02 AM Page 160

hold each other accountable within four rela-tionships (figure 9.5):

• Client power connects service users withproviders.

• Voice connects citizens with politiciansand policymakers through the politicalprocess.

• Compacts connect policymakers throughimplicit or explicit contracts with providersresponsible for services.

• And management connects providerorganizations with frontline across-the-counter providers.

Short route of accountability. In a simplemarket transaction, the buyer holds theseller accountable for the product bought,rewards the seller by repeating business, orpenalizes the seller by choosing anotherprovider. This accountability is “short”because the client can hold the providerdirectly accountable, without any interme-diaries. Small, independent providers inwater and sanitation and their clients areusually in such a market relationship.

In Dar-es-Salaam, Tanzania, a choleraoutbreak in 1996 forced the sewerage andsanitation department to loosen itsmonopoly on cesspit cleaning and allowprivate providers in. There is now anemerging competitive private market forcesspit cleaning—households can choose aprovider based on price and (easy-to-mon-itor) performance. Besides allowing entryand implementing regulations on sewagedisposal, the city’s role has been small.456

But service and market conditions thatautomatically give clients power—throughchoice, ease of monitoring, and marketenforceability—are not always present forinfrastructure services. So the route ofaccountability has to be long.

Long route of accountability. Govern-ments worldwide deem it their responsibil-ity to provide, finance, regulate, and inother ways influence infrastructure services.They do it for two good reasons: marketfailures and equity concerns. First, net-worked infrastructure services exhibit

Drinking water, sanitation, and electricity 161

0.005

0.010

Lima Kampala Bandung Dar es Salaam

Watertrucker

Utility

Bicyclewatervendordeliveringto non-service area

Kiosks

Watervending

Houseconnection

Handcartsdelivering tohomes

Standpipesdrawingwater frommains


Figure 9.4 Alternative sources of water: poor people pay morePrice of water per liter, U.S. dollars

Figure 9.3 Water and sanitation bypoorest and richest fifths

0

Poorest fifth

Richest fifth20

40

60

80PercentAccra

0

20

40

60

80Jakarta

0No water atresidence

20

40

60

80

Sharing toiletswith 10 or more

households

Sao Paulo~


19_WDR_Ch09.qxd 8/14/03 10:02 AM Page 161

economies of scale, or network externali-ties, that make it technically more efficientto have a single distributor of the service. Insanitation the externalities come literallyfrom spillovers. Yes, households in Dar-es-Salaam were willing to pay for improvedsanitation with larger health benefits to thecity. But free-rider problems, where oneperson’s behavior hurts others withimpunity—as in the case of runoff fromopen defecation in many parts of Asia andAfrica—require community or governmentintervention. Second, societies care aboutequity, and governments often redistributeresources—such as a lifeline water sub-sidy—to ensure the minimum equitableservice access that markets cannot.

Network externalities, collective actionproblems, and distributional goals thusprovide powerful reasons for the govern-ment to be involved. The arrangementsthen are no longer primarily between theclient and the provider, and new account-ability relationships become important.The first of these arrangements is voice—citizens delegating to politicians the respon-sibility to ensure the infrastructure ser-vices they want. The second is through thecompacts between policymakers andproviders—to design the service deliveryframework, choose a provider, and ensurethat it meets citizen expectations. Voice andcompacts together become the “long route”of accountability. In Bangladesh the primeminister and her power minister are, inprinciple, accountable to citizens for theperformance of the Power Development

Board, a corporation owned and operatedby government.

The short and long routes of accountabil-ity need to work together. Indeed, even forcesspit services in Dar-es-Salaam, govern-ment regulation was necessary to ensure thatthe small private operators complementedthe public provider and complied withsewage disposal guidelines. Effective solu-tions are likely to be a strategic mixture of theshort and long routes of accountability as asystem in which the clients, the policymaker,and the provider are linked in accountabilityrelationships that make services work forpoor people.

Why infrastructure services fail poorpeople: patronageBecause the family has been without daytimewater for the past decade, the children have neverseen water come out of their home faucets. . . . Thefaucet flows only between midnight and 4 a.m. inmost of Baryo Kapitolyo. MWSS, you know that.Did you care?

Dahli Aspillera, a citizen of Manila, on the eve of the privatization of Manila’s public water agency, Metro-Manila Waterworks

and Sewerage System (MWSS), in 1997

Where water, sanitation, and electricity arepublicly managed, the accountability to cit-izens is achieved when the state ensures thatutilities, boards, and government depart-ments provide efficient and equitable ser-vices for all citizens, including the poor.When the state is unsuccessful and the voicerelationship is not effective, the long routeof accountability has failed.

In 1997 the MWSS was typical of serviceutilities, boards, and government depart-ments that consider politicians and policy-makers as their real clients. Politicians—responding to equity concerns or, morelikely, to short-term political gain—oftenkeep prices for infrastructure services wellbelow those for cost recovery. This makesservice providers dependent on politicallymotivated budget transfers for survival—orwhen transfers are not forthcoming, on ser-vice cutbacks that attract no penalties frompolicymakers.

State-owned water and electricity pro-viders then cease to function as auton-omous service providers.457 They become


Figure 9.5 Accountability in infrastructure services

Voic

eCompact

The state

Politicians Policymakers

Citizens/clients

Client power

Nonpoor Poor

Long route of accountability

Frontline Organizations

Management

Short route

Coalitions/inclusion

Public and private utilitiesand providers

Water, sanitation, other services

19_WDR_Ch09.qxd 8/14/03 10:03 AM Page 162

extensions of policymakers. The policy-maker and provider begin to fuse into onerole. When this happens, policymakers canno longer hold providers accountable fordelivering to all citizens, services deterio-rate, and poor citizens as clients are leftpowerless.

The dynamics of this relationship can beeven more debilitating for poor clients.Over time providers become a strong polit-ical force, influencing the policymaker. Ineffect, providers capture the policymakingprocess, exerting pressures through orga-nized labor or their ability to control servicedelivery for the politician. With deteriorat-ing service levels, policymakers andproviders ration access. This has an impor-tant implication when lumpy investmentsare needed to gain access to services—whether through an electricity grid, a vil-lage water network, or even a stand-alone

system, such as tubewells. Citizens or theirgroups respond to rationed access by sup-porting politicians who favor them as theirclients over politicians who push for uni-versal access. This strengthens the ability ofpoliticians to use patronage. The account-ability linking clients, politicians, policy-makers, and providers is displaced bypatron-client relationships—clientelism—on both legs of the long route of account-ability (box 9.1).458

In such settings, the breakdown in voicefor poor citizens is reinforced by their lossof client power. Dahli Aspillera’s question—did you care?—reflects both a sense that theclient cannot penalize the provider for poorservice and a deeper reality that the longroute of accountability has failed the citizen.If failure of voice is at the root of weak ser-vice delivery in water, sanitation, and elec-tricity, what are the options for reform?


With patronage, the compact between the politicianand provider—the utility or board—is neither trans-parent nor determined by universal client needs.Politicians exert their control by appointing (and dis-missing) company directors and by providing publicsubsidies to finance investments and prop up ailingenterprises. In return for this patronage, water com-panies are often obliged to supply political favors inthe form of excess employment, the depressing oftariffs, political targeting of new investments, andthe distribution of contracts on the basis of politicalcriteria.The consequences: spiraling costs, low ser-

vice quality, and precarious finances.The scarcity ofresources for investment leaves much of the popu-lace without adequate services and forces them torely on expensive or inconvenient alternatives.

The clientelist model broadly describes the poli-tics of urban and rural regional utilities in both waterand electricity sectors. It also applies to local admin-istrations in charge of urban or village-basedservices (funding of piped networks or communitytoilets or even public investment in deep tubewells).

Source: Foster (2002).

B O X 9 . 1 Clientelism in service delivery

Patronage weakening accountability in the citizen-provider chain

Politicalfavors

Operationalsubsidies/

Appointmentof directors

Artificiallydepressed

tariffs

Poorquality ofservice

Politicians

Connected population

Unconnected population

High prices

OverstaffingEmployeesUntendered

contracts ContractorsUtility company

19_WDR_Ch09.qxd 8/14/03 10:03 AM Page 163

Urban water networksWho is the Water Board accountable to?

Question asked of the Managing Director of the Hyderabad Water Board by a consumer,

Hyderabad, September 2002

In cities and towns, where scale economiesprevail, water systems have major networksuppliers—generally a public sector provider,such as the Lagos water board in Nigeria, or asmall municipal water department, as in Cha-pai Nawabganj in Bangladesh. Some of theseproviders belong to local governments—as inthe case of the Johannesburg water utility;some to a state government—as is commonin India; and some—like MWSS in Manila—to central governments. For all, the relevantquestions are whether there is a clear delin-eation of roles between the policymaker andthe provider—and whom the provider isaccountable to, the policymaker or the client?When voice and politics fail, the distinctionbetween the two is blurred, and the provider isaccountable to the policymaker.

Four reform strategies can potentially sep-arate policymakers and providers: decentral-izing assets, using private participation inoperations, charging for services, and relyingon independent providers to give clientschoice. The first two aim to influence com-pacts, the second two to strengthen clientpower. All are politically difficult to imple-ment. That is not surprising, since strongpolitical forces—not technocratic failures—blur the roles of policymakers and providers.The issue is whether these strategies can pro-

vide incentives to remove patronage and com-pensate for the weak voice of poor people.

Strengthening the compact:decentralizing assetsDevolving responsibilities to different tiers ofpolicymakers and separating powers betweenthem can create the right incentives toimprove service delivery. First, by having ser-vice and political boundaries better coincide,decentralization can strengthen voice andaccountability. Second, when the center is incharge of both regulatory and service deliveryresponsibilities, it has few incentives to holditself accountable. Devolving services toanother tier of policymaker triggers incentivesmore compatible with having the center (or anupper-tier government) oversee the regulatoryframework. Finally, devolution creates anopportunity to benchmark performance anduse fiscal resources and reputation as rewardsto support efficient service provision. The con-testability for resources in this context requiresa tier with fiscal capacity and without serviceprovision responsibilities—appropriate forthe center (or a state in a federal system).

Devolving responsibilities to local gov-ernments has had mixed results in waterand sanitation, often leading to the loss ofscale economies, eroding commercial via-bility by excessive fragmentation, and evenconstitutional conflicts between municipal-ities and upper-tier governments.459 Thehistorical experience of industrial countriesoffers lessons for addressing these problems(box 9.2).


In France water assets have historically beendevolved to the commune—the lowest tier ofgovernment. Clusters of communes have inte-grated the industry by delegating water andsanitation services “upward” to private or semi-public companies.The functional boundaries ofthe companies cut across several communes,which continue to own the assets but contractout the management of services.

In the United States water and sanitationassets are also devolved to local governments.Where local governments have been carved upinto small political jurisdictions and individualwater works are impractical, privately ownedcompanies have emerged to provide regionalservices covering several local governments.

Examples include Elizabeth and Hackensack,both in New Jersey.

Interestingly, for France and some areas in theUnited States, the limited capacity of the smallerlocal governments provided the incentives for pri-vate companies to serve clusters of political juris-dictions. In both France and the United States themultijurisdictional coverage prevents the waterprovider from being captured by any one localbody—thus maintaining the separation fromlocal policymakers.

The approach was different in England andthe Netherlands. At the outset of the 20th cen-tury in both countries, oversight and direct pro-vision of water services were in the hands oflocal authorities. In the Netherlands these were

under company structures, mostly owned andrun by municipalities, but many were under pri-vate operation if not ownership. In England thenational government consolidated the localwater systems into regional bodies, movingfrom 1,400 in World War II to 187 in 1974 and 10in the 1980s, all eventually privatized. In theNetherlands, also under central governmentmandate, the municipal companies were con-verted to regional companies to support theexpansion of services to rural areas. But thecompanies remained under the ownership ofmunicipalities and provinces.

Sources: Lorrain (1992); Seidenstat, Haarmeyer, andHakim (2002); Jacobson and Tarr (1996).

B O X 9 . 2 Decentralization and the water industry—in history

19_WDR_Ch09.qxd 8/14/03 10:03 AM Page 164

Fragmentation and the loss of scaleeconomies can be partly addressed by permit-ting interjurisdictional agreements. In theFrench syndicat model, municipal jurisdic-tions can cede the right to provide water andsanitation services to a company jointly ownedby several local authorities. Bolivia’s water lawexplicitly allows for multi-municipal compa-nies. Colombia empowered its regulator toenforce mergers of nonviable local water agen-cies, but ironically exempted the smallest ofthe municipalities that would have benefitedmost from this rule. Brazil’s state companieswere created through voluntary agreementswith municipalities, financed by central funds.

These examples suggest an importantapproach for aligning general decentraliza-tion with sectoral priorities. When authorityis being decentralized, a window usuallyopens for central government to influencethe restructuring of local services. Decentral-ization gives the center the ability to negotiatethe restructuring of devolved assets throughfiscal incentives—say, by deciding to retainthe liabilities while devolving only the assets.

Where devolution has already happened,the center can provide incentives such asfiscal grants to subnational governmentsthat are dependent on milestones of institu-tional reform. Australia’s federal govern-ment provided grants to states to reform thewater sector. The South African govern-ment is also using central fiscal incentives tosupport municipal restructuring and toinfluence reform of urban services, includ-ing water and sanitation. India’s federalgovernment is exploring a similar policyinstrument—the City Challenge Fund—tocreate incentives for general urban reform,including municipal services.

Such fiscal incentives are more effective ifallocated competitively to local tiers of gov-ernment. But this requires information so thecenter can compare the performance of dif-ferent local governments, promoting compe-tition and accountability. It also requires thatthe policy and legal framework enable localgovernments to have the flexibility to reformservice delivery—to form regional compa-nies and use contracting, for example. Coun-tries such as Pakistan and South Africa thathave recently embarked on decentralizationhave adopted such legislation.

The bottom line: upper-tier governmentscan influence the design of compacts at thelocal level through legislation and incentives.However, as demonstrated by experience inLatin America, decentralization processeshave not always been designed with suffi-cient care to allow these kinds of benefits tobe reaped. The success of managing servicereforms during decentralization will dependon whether broader decentralization policiescan ensure that local politicians and policy-makers bear the consequences of policy deci-sions. Ensuring that decentralization canseparate policymakers and providers at thelocal level requires that it also separate rolesand responsibilities of the different tiers ofgovernment (chapter 10). Without that sepa-ration, decentralization may simply transferpatronage to local levels.

Strengthening the compact: using private participation in operationsOver the past decade, private participationhas grown significantly in water, sanitation,and electricity in different forms and acrossmany regions (box 9.3). In general, privateparticipation in infrastructure has beenadvocated for many reasons, includingaccessing management expertise and privateinvestment and introducing incentives in theoperations of infrastructure services. Privateparticipation is also a direct way of separat-ing policymakers and service providersthrough two aspects of the accountabilitychain—compacts and voice.

In the design of compacts, private pro-viders generally require explicit contractsthat define up front the service responsibili-ties of the provider and the policymaker, theregulatory and tariff parameters, and issuesof access by poor households. In addition,the process of contracting private providerscan strengthen the voice channel, particu-larly if advocacy groups and public informa-tion mechanisms are involved in the process.Indeed, service delivery standards and ser-vices for poor people are often explicit in thepolicy debate on private participation inwater and sanitation.

In many industrial countries the involve-ment of the private sector in service deliv-ery enabled governments to develop the


19_WDR_Ch09.qxd 8/14/03 10:04 AM Page 165

Private investment has been far higher in electricity than in water andsanitation. Not surprisingly, the decline in private investment in the latenineties was more pronounced in electricity. In both sectors, the impactof “large deals” and country specific changes are visible—reflected alsoin the geographic concentration—East Asia and Latin America—of pri-

vate investment. Finally, in electricity privatization dominates; in water,management contract and concessions—public ownership—remainsthe norm.

Source: World Bank, PPI Project Database.

B O X 9 . 3 Trends in private participation: water, sewerage, and electricity

Investment commitments in projects with private participation in developing countries, 1990–2001

0

2

4

6

8

10

0

20

40

602001 U.S. dollars (billions) 2001 U.S. dollars (billions)Water and sewerage projects Electricity projects

Cumulative investment, water and sewerage projects (total $40 billion)

Water and sewerage projects by type (total 202 projects) Electricity projects by type (total 832 projects)

Cumulative investment, electricity projects(total $213 billion)

1990 1991 1992 1993 1994 1995 1996 1997 19981999

AguasArgentina

concession

Manila water system concessions

Chileprivatization

Brazil

2000 2001 1990 1991 1992 1993 1994 1995 1996 1997 19981999 2000 2001

All otherdevelopingcountries

Sub-Saharan Africa

1%

Middle East andNorth Africa

0%South Asia1%

Europe andCentral Asia

8%

East Asiaand Pacific

38%

LatinAmericaand the

Caribbean52%

Sub-SaharanAfrica

2%

Middle East andNorth Africa

0%

South Asia10%

Europe andCentral Asia

9%

East Asiaand Pacific

32%

LatinAmericaand the

Caribbean43%

Concessions44%

Management andlease contracts

20%

Greenfieldprojects

28%

Divestitures8%

Divestitures39%

Concessions4% Management and

lease contracts1%

Greenfieldprojects

56%


19_WDR_Ch09.qxd 8/14/03 10:04 AM Page 166

capacity and political setting to regulate,price, and manage water in public and pri-vate contexts (box 9.4). But in today’s devel-oping countries private participation is beingflung into a context of institutional rigidity,not necessarily conducive to the organicgrowth of formal private participation.Using private provision to drive a wedge intopatronage makes managing private partici-pation intensely political—but potentiallypowerful for increasing accountability.

The proof of this potential is already evi-dent. Formal private participation in waterand sanitation has led to greater demand foraccountability—this, despite accounting for asmall part of total investment in water andsanitation. During the 1990s private invest-ment accounted for only 15 percent of totalinvestment in water and sanitation, coveringless than 10 percent of the world’s population.Even in Latin America, where private provi-sion has made the greatest inroads in thewater sector, it only covers 15 percent of thecontinent’s urban population.460 In addition,in contrast to electricity, for example, publicownership and not divestiture of assetsremains the norm in the sector.

Ulitmately, like decentralization, privateprovision offers an opportunity to influencethe relationships of accountability. And likedecentralization, its success depends ondesign and implementation (box 9.5). Experi-ence so far suggests that regulation and infor-mation—two interlinked parts of overall sec-tor reform—are important in successfullyimplementing private sector participation inwater, sanitation, and electricity sectors and inpromoting greater voice in service delivery.

Regulation. A regulatory system in thisReport’s framework is best defined along thedimensions of accountability between the pol-icymaker and the provider—delegation ofresponsibilities and finance, informationabout the performance of the provider, andenforcement (chapter 3). The regulator couldbe responsible for specific elements of theaccountability chain—just providing informa-tion on performance or also ensuring enforce-ment. Sometimes the policymaker is the regu-lator, and sometimes a dedicated third partyhas this responsibility. Sometimes even anassociation of providers can self-regulate.Whichever method is followed, the regulatory


England. In London private companies sup-plied water for more than 400 years with littlegovernment restriction on entry. Companiescompeted against each other, invested inservice and quality innovations, and increasedhousehold connections. By the 19th centuryLondon’s extensive water system helped makethat city “one of the best housed and healthiestcities in Europe, with a death rate lower thanbirth rate by about 1800, at a time when mostEuropean cities were devourers of men.” Ninety-five percent of London residents received pipedsupply from the private companies, and a major-ity had direct home connection.

Technological change led to significantprice competition, industry consolidation, andhigher prices. And the improved water supplyincreased demand for flush toilets, whichcreated problems of sewage removal.Parliament responded with regulation, and by1908 the private system was nationalized. (In the1980s England shifted back to privateprovision.)

Holland. Between 1853 and 1920 the watersector was dominated by private water supply

companies, which were then progressivelytaken over by municipalities and operated aspublic utilities.The amalgamation waspromoted by central regulation and facilitatedby municipal politics. A major motivation was touse the companies to deliver services moreregionally to rural areas. By the time the publicsector took over, principles of economic man-agement of water services had become wellembedded in the political system. Arms-lengthmanagement of public utilities by municipalitiesbecame the norm.

United States. Between 1800 and 1900, U.S.cities experienced a tremendous growth ofwater works. Initially dominated by private own-ers, half of them were public by 1900.The shiftto public ownership emerged because of con-tracting problems between municipalities andcompanies over water for fire fighting.The diffi-culties of establishing contracts when citieswere growing rapidly, and several urban confla-grations, offered opportunities for both privatecompanies and governments to evade perfor-mance targets or force renegotiations ofcontracts. A lack of metering and direct charg-

ing led to conflicts over fiscal transfers frommunicipalities to companies. Not surprisingly,public ownership increased, and with it the pub-lic system inherited the tradition of managingand regulating water as an economic good.

France. Starting with private provision ofwater at the local government level and main-taining it from the mid-1600s onward, Franceevolved toward public ownership and privateprovision through different types ofmanagement and lease contracts.The reasons:scale issues (small local authorities), the historyof the French legal system, and the role of voicein controlling policymakers.The issue of firefighting did not come up in France, perhapsbecause cities were built with vastly differentmaterials and densities.

Sources: Tynan (2002), Schwartz and Maarten(2002), Crocker and Mastens (2002), and Lorrain(1992).

B O X 9 . 4 Private participation—in history

19_WDR_Ch09.qxd 8/14/03 10:05 AM Page 167

process has to separate policymaker andprovider and preserve its own independence.

Organizing regulation: one size does not fitall. Where voice is strong and supported byan effective legal system, the policymakersand the judiciary do the regulating. In France,where the compact for water is betweenmunicipal policymakers and a private com-pany, regulation is done primarily throughmunicipal monitoring of contracts, withsome support from central authorities.

In countries without a tradition of sepa-rating policymakers and providers and withdiscretionary policymaking, credible regula-tion requires a third party—an agency—toset or interpret regulatory rules. Several for-mal safeguards can support the indepen-dence of a regulatory agency from politicalinfluence.461 Some examples: earmarkingfunds for the regulatory agency, hiring stafffrom the market without being restricted bycivil service rules (competence and capacityare important elements of gaining credibilityand independence), ensuring that the hiringand firing of regulators are protected fromthe political interference of the executive andlegislative branches, and not linking theterms of staff to electoral cycles.

A multi-tiered governmental structureoffers additional scope for protecting the inde-pendence of a regulator by placing it at the

national level, or at the state level if policymak-ing and provision are done at the local level.Another option in a multi-tiered governmentis to use local regulation but have the appealsprocess at a different level. In the United Statesthe Constitution provides an overall frame-work for property rights while state regulatorycommissions oversee the operations of pri-vately owned local utilities. Local governmentsregulate public utilities directly.

Regulation and sector reform. The account-ability framework clarifies the conditionsunder which a regulator will be effective insupporting sector reforms. Just as account-ability is blurred if any one of its relationshipsis broken (see chapter 3), the effectiveness ofa regulator is abridged if delegation ofresponsibilities and finance between the stateand the provider is incomplete. That is thecase in the electricity sector in some states inIndia. In other words, an independent regula-tor is needed to enforce the separationbetween policymaker and provider, but if theseparation is not initiated through generalreform to begin with, the regulator may wellbe ineffective. A regulator cannot substitutefor broader sector reforms.

At the same time an effective regulator canhelp sustain sector reform. A recent study ofabout 1,000 concessions in Latin Americashowed that even a moderately well-func-


In the 1990s Argentina embarked on one of thelargest privatization campaigns in the world aspart of a structural reform plan.The programincluded local water companies coveringapproximately 30 percent of the country’smunicipalities. Child mortality fell by 5–7percent in areas that privatized their water ser-vices.The largest gains were seen in the poorestmunicipalities, where child mortality fell by 24percent. Overall, privatization of water servicesprevented approximately 375 deaths of youngchildren each year.

Aggregate data from other sources on thedistribution of new water connection byincome quintile from three countries in LatinAmerica confirm the results of the pro-poorimpact of private sector services. As the datashow, 25–30 percent of the network expansionwas targeted at the lowest 20 percent of theincome profile.

Responding to the need for alternatives forreaching poor people, one of the Manila conces-sionaires has developed a system for water deliv-ery in densely populated, hard-to-reach slumareas. In the Bayan Tubig (“Water for the Commu-nity”) program, the use of appropriate technologi-cal standards, client participation in maintenance,and community-based organizations in interme-diation and mapping of the network reducedwater costs for poor families by up to 25 percent.To increase affordability, the concessionaire hasintroduced an interest-free repayment schemeover a period of 6 to 24 months. Between 1991and 2001, the program provided water connec-tions to more than 50,000 households—thisdespite the fact that the contract of one of theManila concessionaires is under review.

Source: Galiani, Gertler, and Schargrodsky (2002);Water and Sanitation Program (WSP-AF) (2003).

B O X 9 . 5 Private participation in water and sanitation can save poor people’s lives, and money

0Poorest 2 3

Income quintile4 Richest

20

15

10

5

30

25

35

New ConnectionsPercent

BoliviaArgentina

Chile

Distribution of new connections followingprivate sector participation in water andsanitation services

Source: Foster (2002).

19_WDR_Ch09.qxd 8/14/03 10:05 AM Page 168

tioning regulator can temper opportunisticrenegotiations of contracts.462 The study con-cludes that where a regulatory body exists in acountry, the probability of a renegotiation is17 percent; where none exists, the probabilityis 60 percent.

Regulating the public sector. Sector regula-tion is often discussed in the context of pri-vate sector participation. But issues ofmonopoly behavior and service performanceare also relevant for public sector provision—perhaps even more, because the contractsbetween the policymaker and the publicprovider are often not explicit. Independentregulation of public providers is thereforeequally important. But unless publicproviders have operational flexibility and arebrought under explicit compacts—andunless all the relationships of accountabilityare applied—it is not clear how regulation ofthe public providers would have an impacton service standards. In particular, becausemost of the instruments of modern regula-tion are based on financial incentives, in theabsence of user charges regulation of publicproviders would be ineffective. In Chile pub-lic sector regulation was introduced in thecontext of sector reforms, which includedgreater provider autonomy in operations andeconomic pricing of water. This helped cat-alyze regulatory capacity in the public sec-tor—an important asset, now that Chile hasprivatized water services.

The role of information. With private provi-sion more needs to be done to deliver on thedemand for greater voice—informing com-munities about the why and how of privatesector contracting. A public opinion poll inPeru found support for privatization of elec-tricity among only 21 percent of the citizens.But when informed that privatization was tobe undertaken through a transparent processand tariff increases would be regulated, sup-port increased to 60 percent.463 In Manila theconcession process was preceded by a wide-spread public campaign by President Ramos,who convened “Water Summits” to bringtogether different stakeholders.464 In SouthAfrica Johannesburg’s water managementcontract was also undertaken after signifi-cant—and often difficult—consultation with

communities, labor unions, and other interestgroups. Neither process was flawless, but bothopened the door to greater accountability. Anopen process is needed to broaden the partici-pation of communities in the policy debateon private provision—otherwise narrowinterest groups can capture the informationand representation.

Community involvement is also essentialin the regulatory process—but it has not beensufficiently encouraged. A review of urbanwater utilities in Latin America and Africaconcludes that giving consumers little infor-mation about the process of reform and tariffsetting—and limiting their opportunity forcomment before taking regulatory deci-sions—weaken the regulatory process andthe credibility of reform, and make tariffchanges—however justified—difficult toimplement.465

Organizing consumers is, however, not aneasy task. There are major free-rider (andrelated financing) problems in developingcountries that prevent consumers from orga-nizing themselves to a degree where they canbe an articulate voice in the regulatoryprocess. The problem is even more severe forpoor consumers. In industrialized countries,relatively well-developed consumer associa-tions perform this role reasonably effectively.Where competent and effective consumerassociations are absent, the asymmetrybetween consumers and providers becomesmore acute, and the regulator risks being cap-tured by the provider.

Examples exist of regulatory bodies engag-ing communities—especially poor communi-ties—more actively. In Jamaica the regulatorreaches out to communities through localchurches; in some cases in Brazil special con-sultative or advisory bodies have been created;and in Peru regulators have made extensiveuse of the radio to engage and communicatewith communities.466 But these are few exam-ples only—much more needs to be learnedabout how to organize and access communi-ties in the regulation of services.

Managing private participation also re-quires information on how private playersare performing relative to their contractand the performance of other public andprivate providers. This information, whichis critical for regulators, also strengthens the


19_WDR_Ch09.qxd 8/14/03 10:06 AM Page 169

relationship between citizens, politicians, andpolicymakers. For private provision to have acatalytic impact on the sector, information isessential on the performance of both thepublic and private sectors. But too little infor-mation has been available on the perfor-mance of the public sector and through fewcredible sources. Leveling the playing fieldbetween public and private providers—asdiscussed later—and benchmarking theirperformance are essential in getting the bestout of private participation in the sector.

Overall, the impact of private sector partic-ipation is best leveraged within a broaderreform context—greater separation of policy-makers and providers for all public providers;greater participation of communities in theprocess of private participation and in theregulatory framework; and greater use ofbenchmarking of both public and privateproviders. The Australian approach is in-structive. An enabling framework and anational competition law level the playingfield for all public and private providers. Sec-toral legislation provides guidelines for serviceprovision. The central government providesfiscal incentives to support change at the statelevel. A variety of delivery approaches are sup-ported—corporatization (Melbourne), man-agement contract (Adelaide), vertically inte-grated public utility (Sydney), multi-utility(Canberra). Regulation differs between statesand is backed by independent regulatoryagencies as well as benchmarking donethrough an association of water providers.

Strengthening client power:charging for servicesUser charges provide operational autonomyfor the provider, support client power, andelicit greater accountability from the state (box9.6). Without access to enough revenues fromthe clients, service providers depend on thepolicymaker for fiscal resources to maintainservice provision. In addition, if the seller isnot dependent on the buyer for at least somepart of revenues, the provider will have littleincentive to respond to the client. At the sametime, given the politics of water pricing, imple-menting user charges can quickly elicit a con-sumer response—as in Johannesburg, Manila,and very visibly in Cochabamba, Bolivia.

Implementing user charges. Drawing on thepower of user charges to leverage accountabil-ity in service delivery requires, as discussedearlier, effective regulation to address mono-poly provision. But more importantly, the crit-ical policy issue is how to increase tariffs. Thereare two implementation issues: the first is syn-chronizing tariffs with quality improvements,and the second is ensuring that there is a safetynet to safeguard basic affordability.

In many countries, bringing the tariffs tocost-recovery levels would require significantadjustment and rebalancing of tariffs amongresidential, business, and industrial cus-tomers. In Indian cities the charges on resi-dential users are less than a tenth of the oper-ating and maintenance costs. Industrial userspay ten times more but are below the bench-mark for operating and maintenance costs intwo-thirds of the metropolitan cities and 80percent of smaller cities.467 Even if there is awillingness to charge, how can the transitionto prices be managed?

Charging cannot be assessed independentof the broader policy framework and thecredibility of service providers. Policymakersare obviously concerned that services will notimprove enough to justify the price increases.Central to a price increase is what comesfirst—the increases or service improvement?Guinea entered a lease contract for water ser-vices in its major towns and cities in 1989.During the first six years of the contract, thegovernment subsidized a declining share ofthe private operator’s costs while tariffs wereadjusted gradually toward cost recovery,


Treating water as an economic good and charg-ing for services enabled France and the Nether-lands to use private provision to jump-start thesector’s development. In France the private sec-tor remains the major service provider of waterand sanitation services. In the Netherlands thesystem shifted from the private to the publicsector. But in both countries charging users forwater remained the norm, which enabledproviders to sustain service delivery at arm’slength from local government and gave themgreater incentives to be responsive to the needsof the clients.

Sources: Lorrain (1992); Blokland, Braadbaart, andSchwartz (1999).

B O X 9 . 6 Charging forwater—in history

19_WDR_Ch09.qxd 8/14/03 10:06 AM Page 170

which avoided a major tariff shock. Thisjump-started the move to cost recovery andbetter service delivery. It also gave the reformcredibility in a region that had little experi-ence with private provision.468 For variousreasons the lease contract expired in 1999 andwas not renewed, but the pricing strategyremains relevant for other countries.

Similarly, subsidies to poor people couldbe better targeted and designed, whichwould enable user charges to be imple-mented overall. Chile has a nationallyfunded household water subsidy. Colombiauses geographic targeting. South Africa has anational lifeline tariff system that guaranteeseach household 6 kiloliters of water amonth.469 Given the substantial divergencebetween piped water prices and the high costof the inferior alternatives that many of thepoorest are forced to use, there is often astrong case for giving highest priority to con-nection subsidies rather than subsidizing theuse of water by those who already enjoyaccess to the piped network. Connectionsubsidies also have the advantage that theyare easier to target (since lacking access toservice is already a strong indicator ofpoverty) and cheaper to administer (sincerelatively large one-time payments areinvolved). Generally it is more efficient tosubsidize the connection costs for low-income households, but there are alternativeoptions for designing connection and con-sumption charges that benefit poor people.

Ultimately, tariff adjustment and subsidymechanisms are technocratic tools that canbe designed and applied in many ways. Whatis critical is to turn payments for services intoa political tool for reducing patronage andstrengthening client power of poor people.

Strengthening client power: relyingon independent providersAs the example of pit operators in Dar-es-Salaam suggests, small independent pro-viders are a common feature in providingwater and sanitation services across incomegroups. Their organization varies fromhousehold vendors of water, small networkproviders, and private entrepreneurs to coop-eratives. In some cases they are the primarysuppliers, and in others they supplement theformal provider. In some cases they are part

of a competitive market, and in others theyare controlled by a few groups.

Enhancing the role of independentproviders as part of the short route ofaccountability is a key policy challenge. Howcan this be achieved?470 By recognizing inde-pendent providers and giving them legal sta-tus, by ensuring that network providers arenot given exclusive supply, by enablinggreater partnership between formal publicand private network providers and smallindependents, by ensuring that the regulatoryframework for network providers gives theflexibility to enable contracting with inde-pendent providers, by enabling small-scaleprovider associations and working with theseumbrella bodies to introduce appropriatelevels of regulation, and by enabling poorpeople to gain access to multiple independentproviders while keeping their regulationmore focused on health and issues related togroundwater depletion.

Of particular concern is the effect ofbringing in a formal private provider in anarea dominated by independent providers.This issue was not addressed in the design ofthe Cochabamba contract—where the pri-vate provider was given exclusivity rights—and it contributed to the contract’s cancella-tion.471 In reality, if coverage targets aredefined in such a way that they can be metwith the services of small independentproviders, the operator will have an incentiveto encourage their involvement.

Rural areas: network and non-network systemsRural settings are complex in their settlementpatterns, ranging from dense settlements inSouth Asia to dispersed communities in manyAfrican countries. Suppliers include house-hold systems in Bangladesh, water vendors inLaos, and community-managed local pipedwater systems in Ghana. Across all situations,the client-provider link is the norm. Under-standing why the long route of accountabilityis needed to support this client power, andhow this can be done, are the main servicedelivery challenges in rural areas.

Community-managed networksIn countries as diverse as India and Kenya,water boards or engineering departments


19_WDR_Ch09.qxd 8/14/03 10:07 AM Page 171

have traditionally been responsible fordelivering water services to rural communi-ties. Top-down in their approach, with littleskill in community mobilization, andbacked by fiscal support from central gov-ernment, the boards scaled up physicalinvestment. But they had little success inensuring sustainable operations and main-tenance. Indeed, these boards face the sameproblems of state capture inherent in thepatronage model of service delivery.

Given the failures of top-down institu-tions, some countries are shifting to com-munity-managed systems—often supportedby donors, as in India and Ghana. Commu-nities are involved in the design and man-agement of their water systems, paying foroperations and maintenance costs. Govern-ments, generally central governments, pay asignificant part of the capital costs. Donor-funded project management units, backedby not-for-profit organizations, often formthe technical and organizational backboneof these systems.

The client-based model puts the client atthe center of the accountability relation-ship, but many challenges remain in scalingit up.472

• Communities require technical supportin the medium to long run to managewater systems, and donor-funded pro-ject management units are not wellsuited for this.

• Communities pay for current operatingcosts, but replenishing capital invest-ments and covering higher tariffs—topay for rising power costs, for example—are not easily managed through groupcontributions.

• Communities are not homogeneous—problems of exclusion and elite capturecan be the same as in government sys-tems. And different communities mayhave differing abilities to form cohesivegroups.

• Efficient technologies that require scaleeconomies are not selected because ofthe focus on village-level associations.

Supporting client provisionThree approaches—local governments, re-gional utilities, and independent providers—

provide examples of institutional mecha-nisms for supporting community-based sys-tems. They are all “works in progress,” andlearning from them will offer insights onhow to advance rural community-based sys-tems of delivery.

Local governments can form the institu-tional and financial support for expandingcommunity-based systems. With access to atax base, local governments can provideresources to cover periodic capital expendi-ture, provide temporary fiscal support tocommunities to adjust to economic shocks,and facilitate access to technical assistance.Uganda and South Africa provide examplesof arrangements in which local govern-ments are part of a larger fiscal decentral-ization program with own resources andgreater autonomy. Local governments thusstrengthened can support community-based programs. Even in India, where localpanchayats do not have as much autonomy,the relations between local governmentsand user groups are evolving. Where neigh-boring small towns have effective providers,these can be contracted in by rural localgovernments to support their communities.

In Côte d’Ivoire a national utility run bya private partner has responsibilities forurban centers and smaller towns. Thenational utility uses cross subsidies—withthe capital city providing the fiscal sur-plus—to support the smaller urban centers.Expansion of its responsibilities to ruralareas is now being tried. The early lessonshave not been successful but the approach isstill evolving.473

Finally, communities can contract with athird party or an independent provider tomanage local network systems. In Chinaformal cooperatives (rural companies) runon commercial principles with very highcost recovery.474 In several African countriesvillage entrepreneurs manage water systemsunder contract. In East Asia small indepen-dent providers are being organized to takeon operational responsibility on a conces-sion basis. In each case, the process is orga-nized through group consultation andendorsement. While small systems can becontracted by community organizations,villagewide systems may again require thesupport of policymakers at the local level.


19_WDR_Ch09.qxd 8/14/03 10:07 AM Page 172

Self-provision. Households managing wellsand hand pumps are common in large partsof rural Asia and Africa. Nowhere is self-provision more dramatically showcasedthan in Bangladesh, where shallow aquifersand the market provision of hand pumpsenabled households to directly managewater services and replace pathogen-conta-minated surface water with groundwater.Service delivery improved—less waitingtime, no quantity limits, and the conve-nience of household connection. And thehealth impact, which included a decline indiarrhea-related deaths, was remarkable.

Missing was any attempt to monitorwater quality. Finding arsenic in thegroundwater caught everyone by surprise.The government had withdrawn from therural water sector, assuming that access wasnow fully addressed by the private marketand household efforts directly. In addition,in a unitary system of government, therewas no local government to respond to thecrisis. In rural Bangladesh today, a policy-maker is needed to support communities,manage externalities, and understand thetechnological choices for addressing thearsenic crisis (box 9.7). More broadly, for acollective good such as the monitoring ofwater quality, a partnership between clients

and providers will not suffice; policymakersare needed to support communities

SanitationPolicy issues in sanitation need to be dis-cussed in the context of the private and pub-lic goods dimensions of the sector. To theextent it is primarily a client-provider rela-tionship, households invest in sanitationsystems and contract independent providersfor the removal of excreta. To the extent thepublic goods dimensions are dominant, pol-icymakers need to support collective actionto change behavior at the household andcommunity levels, and organize commoninfrastructure for excreta removal.

Access to sanitation services has oftenbeen seen as an issue of subsidizing latrinesand prescribing latrine technology. Thissupply-driven approach, emphasizing thefiscal and engineering aspects of sanitation,has failed. In response, some countries havebeen shifting toward “complete sanita-tion”—focusing on community and house-hold behavior and sanitation practices.475

This involves breaking the fecal-oral chainby encouraging households to changebehavior—shifting away from open defeca-tion, washing hands, keeping food andwater covered, using safe water, focusing on


The arsenic contamination of shallow aquifersmay be undoing the success of rural drinkingwater provision in Bangladesh.While the num-ber of individuals showing symptoms of arsenicpoisoning is still low—despite the high concen-tration of arsenic in the water—between 25 and30 million people may be at risk in the future.

The first response to the crisis bygovernment and many donors was denial.Thiswas followed by an effort to test all watersources and hand pumps.There were varioustechnological and logistical problems—which isnot surprising in view of the fact that arseniccontamination of this scale has not been facedanywhere in the world.These problems werefurther complicated by a lack of coordinationand blurring of roles among government,donors, and nongovernmental organizations(NGOs).

The efforts so far have revealed that surfacewater does not contain arsenic and that not all

aquifers are contaminated. Government, donors,and NGOs are advocating several options: shift-ing to alternative water sources, including somesurface sources; sharing of uncontaminatedtubewells in villages; sinking deep tubewells inpublic areas; and promoting household filteringtechnologies.The latter, if successful, would pre-serve the use of shallow tubewells—decentral-ized, household means of water access—thathave defined the “water miracle” of Bangladesh.In all of this, little effort was made to understandthe preferences of rural households.

A WSP-BRAC (Water and SanitationProgram–Bangladesh Rural Advancement Com-mittee) team undertook a comprehensive sur-vey of household preferences for differentapproaches to arsenic mitigation in selectedareas of rural Bangladesh.The results reveal thatcommunities place a high premium on conve-nience. Unless the alternatives are as convenientas the current hand pumps, the shift to dug

wells, well-sharing, and other mechanism maynot work. Indeed they have not yet beensuccessful as solutions.

Communities strongly indicated apreference and willingness to pay forcentralized, community-based filtering systems,such as local piped-water systems with a centralfiltering point for chemical and biological conta-minants.The piped water network systemsintroduced in the Bogra area by the Rural Devel-opment Academy suggest the potential of suchsystems in Bangladesh.This has been confirmedby preliminary data, which show the cost effec-tiveness of piped water in settlements that have300 or more households. If implementedbroadly, this approach would dramaticallychange the nature of water institutions in ruralBangladesh—a change that communities arewilling to undertake.

Source: Ahmad and others (2002).

B O X 9 . 7 Fighting arsenic by listening to rural communities

19_WDR_Ch09.qxd 8/14/03 10:08 AM Page 173

children’s hygiene behavior and maintain-ing a clean environment. The use ofhygienic latrines is a result of this process ofchanging behavior.

Because the health impact of a house-hold’s sanitation practices is affected notonly by the household’s behavior but also bythe practices of the community, there is acollective action problem. The provider’srole in ensuring information and social sup-port to households through communitystructures becomes critical. Success dependson making people see themselves as a com-munity, where every member’s behavioraffects the other—a daunting challenge andperhaps the reason why sanitation hasalways lagged behind demand for water.

A participatory focus in rural areasBecause communities need to manage sani-tation collectively, innovative participatoryapproaches are required to generatedemand for it, especially in dispersed settle-ments. The shift from open to fixed-pointdefecation may be motivated by health,safety, and privacy concerns—issues ofimportance to women, who bear much ofthe burden of poor sanitation practices. Inthe approach practiced by Village Educa-tion Resource Center (VERC) and Wat-erAid in Bangladesh, an external group trig-gers community-wide recognition of theneed for better sanitation practices. Thecommunity then takes responsibility forself-regulation—motivating households tostrive for complete sanitation. In East andSouth Asia this has even led to innovationsin latrine technology and micro-creditfinancing for investments in latrines andassociated infrastructure.

Subsidies. The community focus alsochanges the approach to latrine subsidies. InBangladesh, villages in the VERC/WaterAidproject did not require any external subsidy.To assist low-income households, higher-income households provided resources.Once communities focused on the need forcollective responsibility, assisting individualhouseholds to reach community goals wasmore readily accepted. In Vietnam the par-ticipatory approach was supported by a sub-sidy targeted at poorer households.

But even if a subsidy is required, the fis-cal contribution could be delivered to thecommunity, rewarding collective action,self-regulation, and the elimination ofopen defecation. Take one of India’slargest states—Maharashtra state, with 97million people. It subsidized latrine con-struction by households below the povertyline only to discover that close to 45 per-cent of the latrines were not being used.So it shifted its subsidy to a competitivescheme (the Gadge Baba scheme) thatrewarded communities for good sanita-tion practices, using an information cam-paign to define the principles of sanitationand publicizing the names of winning vil-lages. Reputation, recognition, and com-munity rewards became the catalyst. Overa short period an estimated 100,000household latrines were built, and forevery rupee of state resources, localspending on sanitation and related infra-structure increased by 35 rupees.

Local compacts. Making the shift to bettersanitation practices is the first objective—but sustaining the shift is equally impor-tant. The local externalities and the need tounderstand and draw on local conditionsand knowledge suggest that local govern-ments are the appropriate policymaker tier.In Vietnam and West Bengal, India, localgovernments have supported communityparticipation and ensured its continuity byfinancing the work of the service provider,usually a not-for-profit organization. InVietnam some local governments have useda program similar to Maharashtra state’sGadge Baba scheme to acknowledge villageand individual achievements.

Responding to demand in urban areasHouseholds in urban settlements with highpopulation densities often show a greaterdemand for better sanitation facilities. Thecondominial systems in São Paulo, Brazil,and the community sanitation systems ofOrangi in Karachi, Pakistan, and Parivartanin Ahmedabad, India, suggest that informalurban communities may be willing to man-age and pay for efficient systems of sanitationand waste disposal. Small independent


19_WDR_Ch09.qxd 8/14/03 10:08 AM Page 174

providers serving households directly, as inDar-es-Salaam, show that urban householdsdo invest in sanitation. So what are theimpediments to expanding these approaches?

The answer may lie outside the realm ofwater and sanitation—and in the regulatorydomain of urban centers. First, the formalrecognition of informal communities by gov-ernments and the provision of some form oftenure have strongly influenced communitywillingness to invest in household infrastruc-ture and to work collectively on communityinfrastructure (La Paz in Bolivia, Ahmedabadin India). Research on garbage collection ininformal settlements in Indonesia providesempirical evidence of the negative relation-ship between incomplete property rights andcommunity investment in local publicgoods.476 It suggests that improving tenuresecurity increases the probability of garbagecollection by 32–44 percent.

Second, in dense urban areas the munici-pal government’s willingness to allocate somepublic land to sanitation systems has enabledcommunities to develop community facili-ties, contracting them with a third party tomaintain and operate them (Pune in India).Use is restricted to the community through amonthly charge collected by the communityand paid to the operator.

Third, municipal laws need to supportflexible standards and ensure that communi-ties and households can make arrangementswith independent providers. Laws that per-mit exclusive service provision need to bereplaced by laws that permit differentapproaches and standards.

A concluding caveat is, however, neces-sary on the discussion about sanitation. His-torical evidence suggests that demand forwater and sanitation follows a sequencing—water first, followed by sanitation and thendemand for waste water treatment. Experi-ence also suggests that this sequencing isinfluenced by many factors of which servicedelivery arrangements in the sector is onlyone. In this context, policymakers mustremain realistic and patient about how farthey can catalyze the demand for sanitationthrough external interventions. Unless em-bedded in a demand-responsive approach,throwing subsidies at latrines will not resolvethe challenge of scaling up sanitation.

ElectricityLike water, electricity has urban and ruralcomponents—and issues of managing gridand off-grid systems. In the grid setting, theissues of separating the policymaker fromthe provider, charging for services, usingprivate providers, and developing effectiveregulatory systems are similar to water net-work issues (box 9.8).477 A key difference isunbundling (rather than decentralizing)services.

For electricity in rural settings, theextension of the grid network provideslessons for managing non-network systemsin water. And the emerging use of off-gridelectricity systems can draw on lessons fromcommunity-managed water in rural set-tings.

Grid systems. The experience from LatinAmerica, Eastern Europe, and South Asiasuggests that unbundling the electricitychain into generation, transmission, and


Unlike water and sanitation coverage, electricity coverage has increased significantly over thepast decade. But like water and sanitation, electricity faces daunting challenges in South Asiaand Africa and rural areas across most regions in the world. And as in the case of water andsanitation, increased electricity coverage does not automatically imply efficient service deliv-ery.The problems of theft, intermittent supply, shared access—captured broadly under theheading of electricity losses—make wires no different from pipes in the context of creatingaccountability in service delivery.

Source: International Energy Agency (2002).

B O X 9 . 8 Are pipes and wires different?

0 25 50 75 100

World

Developingcountries

Middle East

East Asia

Latin America

South Asia

Sub-SaharanAfrica

North Africa

0 5 10 15 20 25 30 35

NigeriaIndiaTogo

KenyaZimbabweCameroon

EritreaAlgeriaSudan

The PhilippinesEgypt

IndonesiaEthiopia

ChinaOECD

Electrification rates by region

Percent Percent

Electricity losses, selected countries

19902000

19_WDR_Ch09.qxd 8/14/03 10:09 AM 5

distribution components is critical toreforms in the sector—but only if the marketis large enough to support multiple electricitygenerators, and hence genuine competi-tion.478 Unbundling provides scope for com-petition in the relevant sectors, primarilygeneration. Separating the components alsocreates scope for getting better informationabout the cost structure of each part of thechain. The competition and the informationadd to client power.

But transmission and distribution func-tions are monopolies, and without effectiveregulation it may be difficult to ensure theseparation of policymaker and provider, andeven reduce the scope to introduce competi-tion in generation. Unless distribution istransferred to different types of ownership, anational or regional government as a soleowner will not have much incentive to sepa-rate its policymaking responsibilities from theoperations of the distribution system.

Privatizing distribution is a common pol-icy approach, but decentralizing electricityassets to local governments is not generallyconsidered. Even where local governmentsown distribution systems—as in SouthAfrica—the policy discussion is about con-solidating into regional distribution systems.This is driven by economies of scale andscope, and perhaps also by policy decisions tocross-subsidize from urban to rural settingsand to keep the cross subsidy in the sector.Interestingly, in Mumbai and Kolkata, India,where electricity is under local governments,electricity provision has long been under pri-vate operation. Even in Delhi—in effect acity-state—power distribution is now pri-vate.

Rural grid. The extension of the grid intorural areas offer insights for rural water andoff-grid electricity with regard to reestablish-ing the relationship between policymakerand service providers. A model of ruralcooperatives has emerged in the UnitedStates and is being adapted in Bangladeshand the Philippines. A regional or nationalprovider organization contracts with com-munity cooperatives to be village-level dis-tributors. In Bangladesh the Rural ElectricityBoard (REB) supports the village coopera-tives through technical assistance and fiscal

transfers for a part of the capital costs. Thiswould be similar to using regional water util-ities to support community-managed watersystems. Importantly, the owner of the distri-bution is not the policymaker but the clients.Unbundling the national REB into regionalREBs, with some form of benchmarking,could support the clients in breaking a possi-ble monopolistic relationship between theREB and the cooperatives.

Rural off-grid. Rural provider organiza-tions—or local governments—can also sup-port off-grid systems in villages, in manycases using renewable energy to generatepower. Donors have traditionally advocatedsolar household systems—not unlike thetechnology push in latrines. But today’srenewable systems can support villagewidegrids—similar to villagewide piped water sys-tems—to provide AC electricity for house-hold appliances of various types. Dependingon local conditions the systems can also bewind-powered, solar, tidal, bio-gas, or hybrid,with fossil-fueled generators as backup.

Moving the reform agendaforwardIndia is revolting and the Thames stinks.

Slogan in London, 1857The result: Chadwick and the sanitary

revolution in the United Kingdom.479

Given the weak voice relationship betweencitizens and politicians in the water, sanita-tion, and electricity sectors, deep institutionalreform often comes from broader stresses inthe economic, political, and institutionalmachinery of a country. In London pollutionwas such a cause. In Johannesburg the city’sbankruptcy was the impetus. In cities inAfrica and Latin America a core impulse forreform of urban water and sanitation is thecombination of sector problems and amacroeconomic crisis.480

Society’s view of economic development isalso important. In Australia, Chile, and Peru,growth-driven economic development strat-egy provided the impetus for improving theperformance of water and power markets. Sothe possibilities for sector reform seem great-est when there is a confluence of natural chal-lenge, fiscal crisis, and institutional reform-


19_WDR_Ch09.qxd 8/14/03 10:09 AM Page 176

mindedness.479 Opportunities for reform maywell arrive only by chance, when broaderchanges in turn catalyze sectoral reforms.What are the potential interim measures? Canincremental change be strategic?

For urban networks, change will requireseparating the delivery functions from thoseof benchmarking and regulation. Keepingthe latter with an upper-tier government—central or regional—while dispersing own-ership of water and sanitation assets tolower-tier governments and the private sec-tor could create this separation. Withoutownership responsibilities, the upper-tierpolicymaker would have greater incentivesto use fiscal instruments, benchmarking,and regulation to promote improvements inservice provision. Such incentives are lessinherent in a model where the regulator,provider, and owner are one and the same.Charging users for services strengthens thisseparation by directly involving clients inthe service chain through the short route ofaccountability.

Where the introduction of private sectorparticipation is tempered by politics or otherfactors, strategic change may have to comefirst through changes in ownership and rela-tionships of accountability between tiers ofthe public sector. Interestingly, the history ofsome industrialized countries suggests thatlocal ownership can trigger a more crediblepath to private sector participation, especiallyif local governments are effective in strength-ening voice.

Where local governments exist and waterand sanitation services have been devolved tolocal governments, the challenge of improvedservice delivery would lie in making decen-tralization work. Where local governmentsdo not exist, the lever of decentralized owner-ship would be lost, but benchmarking andregulation of the public sector would remain.But for such a strategy to be effective, charg-ing for water would become even more criti-cal. It would enable providers to achievesome independence in operations, but moreimportantly it would give clients a role in sus-

taining the separation and ensuring that theregulation of public providers is effective. Inthis context, introducing private players in afew of the utilities would enhance the effec-tiveness of benchmarking the publicproviders.

Where this broader approach of makingservices work for all is not possible, a targetedapproach for serving poor people using smallindependent providers is still an option.Indeed, increasingly independent providersmay, at the margin, emerge as a critical leverfor making services work for poor people.

For rural systems—community-managedsystems and self-provision—the challenge isto seek mechanisms for the policymaker tosupport client power, using local govern-ments, regional utilities, and independentproviders. This is similar to the model of therural electricity cooperatives supported by aprovider organization that provides a techni-cal and fiscal hub. Where local governmentsprovide this hub, the voice channel is direct;where utilities are the support mechanism,the voice channel is indirect. Where theseoptions are not possible, the approach—however unsatisfactory—of targeted com-munity projects remains.

For sanitation, the focus is on collectiveaction—to change behavior and mobilizecommunities to invest in community infra-structure. To support this, compacts betweenpolicymakers and NGOs may be moreappropriate. In urban areas, where greaterdemand for sanitation services may exist,policymakers can support client power byallowing independent providers to functionand by supporting tenure in informal settle-ments. A more incremental version would besimilar to that in the rural water sector—witha public provider organization supportingNGO delivery in targeted areas.

But if the failure of voice is why infra-structure services have failed poor people,targeted intervention cannot form the basisof institutional reform. Reforming the rela-tionships of accountability would remain thepolicy challenge.


19_WDR_Ch09.qxd 8/14/03 10:10 AM Page 177

They slashed capital and operating budgets,and even expenditures needed to maintainminimum service levels. They froze posts,causing huge increases in workloads asdespairing officials began to drift away. Andthey began to explore public-private part-nerships.

The city of gold—iGoli 2002The new city management team realizedthat Johannesburg needed a new system ofaccountability for service delivery within adramatically different institutional archi-tecture. To address fragmentation and thesevere moral hazard, the city had to bereunified. Political debate focused on twomodels of metropolitan coordination:

• Defining more clearly the rules of bud-geting, fiscal transfers, and service deliv-ery between the metropolitan andmunicipal tiers, strengthening both.

• Creating a one-tier metropolitan gov-ernment.484

Johannesburg chose a hybrid. It central-ized political authority, treasury manage-ment, and spatial planning under one met-ropolitan government. But it organizedservice delivery through decentralizedstructures. This meant merging five sepa-rate councils into one overarching munici-pality, creating integrated service deliverystructures with new incentives.

Accountability in service deliveryUnder one metropolitan council, iGoli 2002split the institution for policy formulationand regulation from the institutions forimplementation. On one side, a core admin-istration remained responsible for strategicplanning, contract administration, and suchcorporate services as finance, planning, andcommunication. On the other, two sets ofoperating entities were established: 11 newregional administrations for libraries,health, recreation, and other communityservices; and financially ring-fenced, semi-

By most developing world standardsJohannesburg is not a poor city. Butit faces serious development and

service delivery challenges. Apartheid madesure that exclusive white suburbs were wellserviced, forcing black residents intosprawling underdeveloped slums. Poverty,unemployment, and homelessness are allworsened by the deeper problem ofinequality.

The Johannesburg Metropolitan Munici-pality was democratically elected in 1995 toaddress the service imbalances. It quicklyfound itself in a fiscal and institutional crisis.

Johannesburg was not one institutionbut five, with an overarching MetropolitanCouncil and four primary-level councils.Each could decide its priorities and approveits budget. But responsibilities for key ser-vices were split between the two levels, andthe operating budgets of the councils had tobalance only in aggregate. That meant eachcouncil could blissfully spend on theassumption that its shortfalls would be off-set by surpluses in another.

The arrangement was a recipe for disas-ter. Each municipality went on a spendingspree, and ambitious infrastructure planswere rolled out without the finance. Deteri-orating revenues—due to a service-paymentboycott culture left over from anti-apartheid struggles, poverty, and poor creditcontrol—made the situation worse. The citywas forced to delve into its reserves, butthese could go only so far, and by late 1997major creditors could no longer be paid. Atthe peak of the crisis, the city had an operat-ing deficit of R314 million.

Johannesburg was in serious trouble.Having decentralized responsibilities, thenational government followed the intergov-ernmental rules and would not bail the cityout. So Johannesburg had to dig itself outof its own crisis.

Two years of harsh cutbacks followed.Blaming officials for the crisis, politicianstook a much tighter rein over day-to-daydecisions, ending management discretion.

independent, single-purpose entities tooverhaul larger municipal services.

These operating entities were the majorinnovation of iGoli 2002.

• Three utilities were established for usercharge–based services—water and sani-tation, electricity, and waste manage-ment.

• Two agencies were established—forparks and cemeteries, and for roads andstorm water—where expenditure wouldstill have to be covered by tax revenue.

• Smaller corporatized units were set upfor facilities like the zoo and the civictheater.

All were established as new companies,with the council as sole shareholder.

Two key units would guide and overseethe new entities: a corporate planningunit to do citywide strategic planning, anda contract management unit to regulatethe operating utilities through a range of new instruments, including licensingagreements and annual service levelagreements.

One size does not fit allSince the operating entities are not boundby overarching administrative rules, theyhave scope to differentiate. Each could setup different management structures,reporting lines, delegations, job descrip-tions, performance management systems,and operating procedures. Each could con-figure its internal accountability to suit aspecific service delivery environment.Three examples:

• The water and sanitation departmentswere merged into one department andunder the Company’s Law convertedinto a city-owned utility with a board ofdirectors. The assets and workers of thedepartments were transferred to theutility, which was put under a five-yearmanagement contract with a privatecompany.

Accountability in city servicesIn 1999 the Transformation Lekgotla, the political body directed to address the financial and institutional crisis of Johan-nesburg, South Africa, appointed a new city management team. The team’s task was clear: not to fix street lights but to fixthe institutions that fix street lights.482 The solution was a three-year plan—“iGoli 2002”—to reconfigure city services.483

Johannesburgs p o t l i g h t o n

20_pgs 178-179_Ch09Spot.qxd 8/14/03 3:39 PM Page 178

Spotlight on Johannesburg 179

also apparent in the city’s financial stand-ing, with dramatic improvement in bothoperating and capital budgets (figure 1).

Engaging other stakeholdersLabor: Despite protracted negotiationswith organized labor, iGoli 2002 did not getits endorsement. According to labor groups,the city’s crisis was not a result of a failureof institutional design. Instead it was aresult of “a lack of skills and experience,and management’s unwillingness to[establish] functional organizations and . . .financially unsound decisions.”485

National government: The team ne-gotiated a R500 million restructuring grantwith the National Treasury to support iGoli2002 in exchange for a commitment totimely and steadfast implementation of itskey elements. It is a key accountabilitymechanism between the national and citygovernments and has become an incentivescheme to catalyze citywide restructuringthroughout the country.

Capital markets: On the strength of thereforms, management sought a new creditrating, aiming to win back the confidenceof the city’s banking community. As the cityshifted from a large deficit to a balancedbudget, capital expenditure financed by themarkets went from R300 million to wellover R1 billion in two years.

• The roads department was convertedinto a city-owned agency with a profes-sional board and divided into twodepartments—for planning and forcontracts. The contracts departmentoperated against specific outcomes setby the planning department, with thethreat that failing to meet benchmarkscould lead to contracting tasks out tothe private sector.

• The gas company was sold to the privatesector.

The reforms gave operating entitiesmanagement independence. For example,salaries have been adjusted to attract top-flight skills, and new systems have beenprocured for everything from humanresource management to remote water-pressure metering—increasing productiv-ity and service efficiency. And they haveintroduced innovative staff developmentprograms and performance-linked payschemes.

The entities operate at arm’s length fromthe council, but accountability has beenstrengthened because the primary mecha-nism is no longer the impossible-to-digestcommittee report on everyday operationalmatters. Now councilors focus on strategicoversight, and officials are responsible foroutcomes clearly defined in service-levelagreements. Reporting goes through struc-tured channels, either to the contract man-agement unit or to company boards ofdirectors, which include external specialistscapable of probing service results.

The operating entities have also set upuser forums allowing communities to com-municate needs, raise complaints, and evenparticipate actively in service provision.Officials are much more sensitive to ever-changing service delivery challenges.

These management improvements arealready translating into better service deliv-ery. Waste collection has been extended topoorer neighborhoods for the first time.Fleets of new buses now serve outlyingcommunities. In addition, expenditure onwater infrastructure has increased andwater services have expanded. Results are

Risks and prospects Will Johannesburg maintain the separationbetween policymaking, providers, and reg-ulators? The roles of client and contractorare still evolving. Some implementationcapacity remains within the core adminis-tration. As in the past, managers occasion-ally get hauled into councilors’ offices toexplain their actions. There are also unre-solved governance debates, with the councilarguing for a greater councilor representa-tion on the boards of operating entities.

Five factors will be critical in sustainingthe commitment to the principles of iGoli2002:

• Keeping the monitoring and regulatoryunit of the operating entities within thecity administration; they are not legallyand administratively independent.

• Maintaining the contract managementunit’s operational autonomy and capac-ity—and thus the independence of theoperating entities.

• Benchmarking service delivery stan-dards, monitoring these over time, andmaking the information available.

• Ensuring that fiscal and financial decen-tralization remains binding. Municipali-ties relying primarily on their own rev-enue sources to fulfill their democraticduties without national guarantees aremore likely to be accountable to theircitizens. The current intergovernmentalsystem has devolved authority andaccountability to the cities; this needs toremain.

• Both councilors and officials consistentlyadhering to a clear, courageous, and far-sighted strategy. Sustaining momentumwill require greater citizen voice at all lev-els. The decentralized operating entitiesand the administrative regions havemechanisms for engaging citizens. Usingthem will be critical for sustaining iGoli2002.

–300

1995 1996 1997 1998 1999 2000 2001 2002

500

–50–100–150–200–250

–350

100150

Millions of rands

Source: Allan, Gotz, and Joseph (2001).

Figure 1 Getting back to an operatingsurplus—thanks to iGoli 2002

20_pgs 178-179_Ch09Spot.qxd 8/14/03 3:39 PM Page 179

drinking water, sanitation, 19 wdr ch09.qxd 8/14/03 10:01 am …. drinking... · 2016-11-22 ·...

Documents