5 Sanitation and clean water

For many people in developing countries water supply, sanitation, and solid wastes are the most important of allenvironmental problems. More than 2 million deaths from diarrhea alone could be avoided each year if all peoplehad reasonable water and sanitation services. And large economic and environmental costs are incurred in tryingto compensate for poor-quality services.

This chapterargues that large gainsin environmental quality, health, equity, and direct economic returnscan be realized by adopting an approach that comprises four key elements:

Managing water resources better, taking account of economic efficiency and environmental sustainability

Providing, at full cost, those 'private" services that people want and are willing to pay for (including watersupply and the collection of human excreta, wastewater, and solid wastes)

Using scarce public funds only for those services (specifically, treatment and disposal of human excreta,wastewater, and solid wastes) that provide wider communal benefits

Developing flexible and responsive institutional mechanisms for providing these services, with a larger rolefor community organizations and the private sector.

Although the provision of clean water and sani-tation is often omitted from the list of priorityenvironmental challenges, in many parts of thedeveloping world it ranks at the top. Two environ-mental issues are involved: the costs to humanhealth and productivity of polluted water and in-adequate sanitation and the stresses placed on wa-ter resources by rapidly growing human demandsfor water. This chapter argues that to address thefirst problem, the second must be tackled as well.This will require better management and more effi-cient use of water. It may mean that agriculturewill have to do more with less water (as discussedin Chapter 7), and it will certainly demand a shiftin how sanitation and water supply services areprovidedthe main theme of this chapter.

Water supply and sanitation as environmentalpriorities

Inadequate sanitation is a major cause of the deg-radation of the quality of groundwater and surfacewater described in Chapter 2. Economic growthleads to larger discharges of wastewater and solidwastes per capita. Inadequate investments in

98

waste collection and disposal mean that largequantities of waste enter both groundwater andsurface water. Groundwater contamination is lessvisible but often more serious because it can takedecades for polluted aquifers to cleanse them-selves and because large numbers of people drinkuntreated groundwater.

More environmental damage occurs when peo-ple try to compensate for inadequate provision.The lack or unreliability of piped water causeshouseholds to sink their own wells, which oftenleads to overpumping and depletion. In cities suchas Jakarta, where almost two-thirds of the popula-tion relies on groundwater, the water table has de-clined dramatically since the 1970s. In coastal areasthis can cause saline intrusion, sometimes render-ing the water permanently unfit for consumption.In, for example, Bangkok excessive pumping hasalso led to subsidence, cracked pavements, brokenwater and sewerage pipes, intrusion of seawater,and flooding.

Inadequate water supply also prompts people toboil water, thus using energy. The practice is espe-cially common in Asia. In Jakarta more than $50million is spent each year by households for this

purposean amount equal to 1 percent of thecity's GDP. Investments in water supply can there-fore reduce fuelwood consumption and airpollution.

Effects on health

The health benefits from better water and sanita-tion, as noted in Chapter 2, are large. When ser-vices were improved in the industrial countries inthe nineteenth and twentieth centuries, the impacton health was revolutionary. Life expectancy inFrench cities, for example, increased from about 32years in 1850 to about 45 years in 1900, with thetiming of changes corresponding closely to im-provements in water supply and wastewater dis-posal (Figure 5.1). Today, adequate water and sani-tation services are just as vital: diarrheal deathrates are typically about 60 percent lower amongchildren in households with adequate facilitiesthan among those in households without such fa-cilities. Box 5.1 describes the improvements thatare critical for better health.

Effects on productivity

Improved environmental sanitation has economicbenefits. Consider the case of sewage collection inSantiago, Chile. The principal justification for in-vestments was the need to reduce the extraordi-

Health gains follow investments in waterand sewerage

Figure 5.1 Life expectancy and improvementsin water supply and sanitation in selectedFrench cities, 1820-1900

Improvements made

Life Lyon Paris Marseillesexpectancy atbirth (years)

50

45

40

35

30

25

4, 4,4,4,4,4,4,4,

1820 1830 1840 1850 1860 1870 1880 1890 1900

Source: Briscoe 1987.

Box 5.1 Specific investments that matter for

The potential health benefits from improved water andsanitation services are huge. What improvements mustbe made to secure these benefits?

Water quality. Contrary to common belief, contam-ination of water in the home is relatively unimportant.What matters is whether the water coming out of thetap or pump is contaminated. In most developingcountries the imperative is to get from 'bad" quality(say, more than 1,000 fecal coliforms per 100 milliliters)to "moderate" quality (less than 10 fecal coliforms per100 milliliters), not necessarily to meet the stringentquality standards of industrial countries.

Water availability. As long as families have to go outof the yard to collect water, the quantities used willremain low (typically between 15 liters and 30 liters percapita per day). The use of water for personal hygieneusually increases only when availability rises to about

health

50 liters per capita per day and generally depends ongetting the water delivered to the yard or house.

Excreta disposal. It is necessary to distinguishamong the effects on the household and on the neigh-borhood. For the household, the health impacts of im-proved sanitation facilities depend only on getting theexcreta out of the house and are thus similar whetherfamily members use an improved pit latrine, a cesspooloverflowing into a street drain, or a conventional sew-erage system. For the neighborhood, the key is theremoval of excreta, a task done well by a wide range oftechnologies but badly by many commonly used sys-tems (such as nightsoil collection and unemptied septictanks). Because all the fecal-oral transmission routesare much more important when people live in closeproximity to each other, the ill effects of poor environ-mental sanitation are greatest in high-density urbansettlements.

99

narily high incidence of typhoid fever in the city. Asecondary motive was to maintain access to themarkets of industrial countries for Chile's increas-ingly important exports of fruit and vegetables. Toensure the sanitary quality of these exports, it wasessential to stop using raw wastewater in theirproduction. In the light of the current cholera epi-demic in Latin America, this reasoning was pre-scient. In just the first ten weeks of the choleraepidemic in Peru, losses from reduced agriculturalexports and tourism were estimated at $1 billionmore than three times the amount that the countryhad invested in water supply and sanitation ser-vices during the 1980s.

Improved access to water and sanitation alsoyields direct economic benefits. For many ruralpeople, obtaining water is time-consuming andheavy work, taking up to 15 percent of women'stime. Improvement projects have reduced the timesubstantially. In a village on the Mueda Plateau inMozambique, for instance, the average time thatwomen spent collecting water was reduced from120 to 25 minutes a day. Family well-being wasthus improved, as the time saved could be used tocultivate crops, tend a home garden, trade in themarket, keep small livestock, care for children, oreven rest. Because users clearly perceive thesetime savings, they are willing to pay substantialamounts (as discussed below) for easier access.

In the absence of formal services, people have toprovide their own services, often at high cost. InJakarta, for instance, about 800,000 householdshave installed septic tanks, each costing severalhundred dollars (not counting the cost of theland). And in many cities and towns large num-bers of people buy water from vendors. A reviewof vending in sixteen cities shows that the unit costof vended water is always much higher than thatof water from a piped city supplyfrom 4 to 100times higher, with a median of about 12. The situa-tion in Lima is typical; although a poor family usesonly one-sixth as much water as a middle-classfamily, its monthly water bill is three times aslarge. Consequently, in the slums around manycities water costs the poor a large part of house-hold income-18 percent in Onitsha, Nigeria, and20 percent in Port-au-Prince, for example.

The economic costs of compensating for unreli-able servicesby building in-house storage facili-ties, sinking wells, or installing booster pumps(which can draw contaminated groundwater intothe water distribution system)are substantial. InTegucigalpa, for example, the sum of such invest-ments is so large that it would be enough to double

100

the number of deep wells providing water to thecity. And the costs of compensating for poor waterquality are great, too. In Bangladesh boiling drink-ing water would take 11 percent of the income of afamily in the lowest quartile. With the outbreak ofcholera in Peru the Ministry of Health has urgedall residents to boil drinking water for ten minutes.The cost of doing so would amount to 29 percentof the average household income in a squattersettlement.

What needs to be done?

Investments in sanitation and water offer high eco-nomic, social, and environmental returns. Univer-sal provision of these services should and couldbecome a reality in the coming generation. But thenext four decades will see urban populations indeveloping countries rise threefold and domesticdemand for water increase fivefold. Current ap-proaches will not meet these demands, and thereis a real possibility that the numbers unservedcould rise substantially, even while aquifers aredepleted and rivers degraded. The remainder ofthis chapter discusses four key policy changes thatneed to be made.

Managing water resources better

When there was little competition for water, it was(correctly) used in large quantities for activities inwhich the value of a unit of water was relativelylow. In many countries irrigated agriculture be-came the dominant "high-volume, low-value"user. Today about 73 percent of all water with-drawals (and higher proportions of consumptiveuse) are for irrigation. This share is even higher inlow-income countries, as shown in Table 5.1. Inmost countries this water is provided at heavilysubsidized prices, with users seldom paying morethan 10 percent of operating costs.

As demand by households, industries, andfarmers increases, governments find it hard to

Table 5.1 Sectoral water withdrawals,by country income group

Annualwithdrawals

per capita Withdrawals, by sector (percent)

Income group (cubic meters) Domestic Industry Agriculture

Low-income 386 4 5 91

Middle-income 453 13 18 69

High-income 1,167 14 47 39

Source: World Resources Institute 1990.

change existing arrangements. The allocation ofwater in all countries is a complex issue and isgoverned by legal and cultural traditions. Userstypically have well-established rights. Reallocationis a contentious and ponderous process that gener-ally responds to changes in demand only withlong lags. Even though agricultural use of waterhas the lowest value per cubic meter, there isstrong political opposition to diverting water fromagriculture to other sectors. The result is that inmany countries, industrial and developing alike,large volumes of water are used in irrigated agri-culture, adding little economic value, while citiesand industries, which would gladly pay more,cannot get enough.

This mismatch is most striking in the areasaround large cities. In the western United States,for example, farmers in Arizona pay less than 1cent for a cubic meter of water, while residents ofthe city of Phoenix pay about 25 cents. In the in-dustrial heartland of China around Beijing andTianjin 65 percent of water is used relatively ineffi-ciently for low-value irrigation, while huge expen-ditures are contemplated to bring water from otherriver basins to the cities.

Paradoxically, there is good news in these dis-tortions. Their very size indicates that urban short-ages could be met with only modest reallocation.In Arizona, for instance, the purchase of the waterrights from just one farm is sufficient to providewater for tens of thousands of urban dwellers. Be-cause of the low value of water in irrigated agricul-ture, the loss of this marginal water has little over-all effect on farm output. To help transfers, newmarket-driven methods for reallocation have beendeveloped. When a recent drought dangerouslyreduced available water, the state of California setup a voluntary "water bank" that purchased wa-ter from farmers and sold it to urban areas. Thefarmers made a profit by selling the water for morethan it was worth to them, while the cities gotwater at a cost well below that of other sources ofsupply.

In developing countries, too, a start is beingmade in applying innovative methods for manag-ing water resources. China's State Science andTechnology Commission found that the economicrate of return to a cubic meter of water used foragriculture was less than 10 percent of the returnto municipal and industrial users. Once agri-cultural and urban users accepted that they had tolook at water as an economic commodity with aprice, progressincluding reallocationwas pos-sible. And Jakarta has been reasonably successful

in reducing the overpumping of its aquifers by reg-istering groundwater users (especially commercialand industrial establishments) and by introducinga groundwater levy.

The striking features of these "market-based"reallocation methods are that they are voluntary,they yield economic benefits for both buyers andsellers, they reduce the environmental problemscaused by profligate use of water in irrigation, andthey lessen the need for more dams.

Without effective management of water re-sources, the cost of supplying water to cities willcontinue to rise. The most dramatic examples willbe in large and growing urban areas. In MexicoCity, where much water is used for irrigation, thecity has to contemplate pumping water over anelevation exceeding 1,000 meters into the Valley ofMexico; in Lima upstream pollution has increasedtreatment costs by about 30 percent; in Shanghaiwater intakes have already been moved upstreammore than 40 kilometers at a cost of about $300million; and in Amman the most recent works in-volve pumping water up 1,200 meters from a siteabout 40 kilometers from the city. A recent analysisof the costs of raw water for urban areas in WorldBank-financed projects (Figure 5.2) shows that theunit cost of water would more than doubleand insome cases more than tripleunder a new waterdevelopment project.

Industries and households also need to be givenincentives to use water efficiently. Cities, likefarmers, have tended to take demand as given andto see as their task increasing supplies to meet it.As was the case with energy twenty years ago,little attention is paid to conservation and demandmanagement in the water sector. This is both eco-nomically and environmentally unsound. Con-sider the case of Washington, D.C. In the 1960s theU.S. government concluded that sixteen dams andmore than $400 million were required to meet thewater needs of the metropolitan area. Because ofresistance from environmentalists to the construc-tion of the dams, the plan had to be reconsidered.Eventually the number of dams was reduced toone and the total cost of the scheme to $30 million.The key changes were a revised plan for managingdemand during droughts and more efficient oper-ating rules. This illustrates once again that bettereconomics and a better environment arecompatible.

Experience in industrial and developing coun-tries alike shows the potential for using watermore cost-effectively in industry. In the UnitedStates withdrawals of fresh water by manufactur-

101

Costs are high and are rising rapidly

Figure 5.2 Supplying water to urbanareas: current cost and projected future cost(1988 dollars per cubic meter of water)

Future cost

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Current cost

Future cost is three times current cost

Future cost is twice current cost

Future cost equals current cost

Note: Cost excludes treatment and distribution. Currentcost refers to cost at the time data were gathered. Futurecost is a projection of cost under a new water developmentproject.Source: World Bank data.

ing industries are expected to be 62 percent less in2000 than in 1977, primarily because of the in-creased costs industries have to pay for disposingof industrial wastewater. In São Paulo, Brazil, theimposition of effluent charges induced three in-dustrial plants to reduce their water demand bybetween 42 and 62 percent. Figure 5.3 shows howin Beijing a variety of conservation measures inindustries and households could release largequantities of water at a substantially lower unitcost than the cost under the next supply augmen-tation project.

102

A particularly important conservation alterna-tive is reclamation of wastewater. Reclamation ofwater for urban, industrial, and agricultural use isattractive both for improving the environment andfor reducing the costs of water supply. Reclaimedwastewater has been used for many years forflushing toilets in residential and commercialbuildings in Japan and Singapore. A recent recla-mation scheme in the Vallejo area of Mexico City(Box 5.2) illustrates the great potential, both eco-nomic and environmental, of wastewater reuseand, to anticipate a theme developed below, thescope for the private sector.

At present, in most countries management ofwater resources is fragmented (industrial users,for example, do not have to take account of thecosts that their use and pollution of water imposes

Box 5.2 Environmental improvement,management of water resources,and the private sector in Mexico

In 1989, faced with rising water prices and poten-tial water shortages, a group of companies in theVallejo area of Mexico City sought an alternativeto water supplied by the public agency. At aboutthe same time, the Mexican government decidedto involve the private sector in water supply andwastewater treatment.

The industrialists realized that if sewage flowscould be adequately treated, this could provide acost-effective and reliable source of industrial wa-ter (and, incidentally, could improve the environ-ment by treating wastes and reducing the need fornew water supplies). Twenty-six Vallejo com-panies organized a new for-profit firm, Aguas In-dustriales de Vallejo (AIV), to rehabilitate an oldmunicipal wastewater treatment plant. Eachshareholder company contributed equity on thebasis of its water requirements, with total equityamounting to $900,000.

AIV operates the plant under a ten-year conces-sion from the government. The plant now pro-vides 60 liters per second to shareholders and 30liters per second to the government as paymentfor the concession. The concession agreementgives AIV the right to withdraw up to 200 litersper second of wastewater from the municipaltrunk sewer. AIV plans to double the plant's ca-pacity within five years at an estimated cost of$1.5 million. The firm provides treated water toshareholder companies at a price equivalent to 75percent of the water tariff charged by the govern-ment (currently, $0.95 per cubic meter).

Conservation can be cheaper than new supplies

Figure 5.3 Conserving water as an alternative to expanding supply in Beijing(discounted cost in dollars per cubic meter)

Domestic uses(total savings: 15 percent of

current domestic consumption)

0.12

0.10

0.08

0.06

0.04

0.02

0.00

o Improved efficiency in public facilitieso Leakage reduction programO Recycling of cooling water used in

air conditioning

Installation of water-efficient flush toilets

Source: Hufschmidt and others 1987.

Cost of next water development project

Industry(total savings: 33 percent of

current industrial consumption)

0 Increased recycling of cooling waterin manufacturing

0 Recycling of cooling water frompower plants

0 Wastewater recycling

on domestic users downstream) and is done by"command and control" (most allocations are setby administrative fiat). The challenge is to replacethis system with one that recognizes the unitarynature of the resource and its economic value andthat relies heavily on prices and other incentives toencourage efficient use of water.

Providing services that people want and arewilling to pay for

During the United Nations Drinking Water andSanitation Decade of the 1980s, coverage increased(see Chapter 2). But about 1 billion people still lackan adequate water supply, and about 1.7 billionpeople do not have adequate sanitation facilities.The quality of service often remains poor. In LatinAmerica, for example, levels of leakage and pipebreakage are, respectively, four times and twenty

times higher than is normal in industrial countries.In Lima 70 percent of the water distribution dis-tricts provide inadequate water pressure. In Mex-ico 20 percent of the water supply systems haveunreliable chlorination facilities.

What has been done

Developing countries cannot afford to provide allpeople with in-house piped water and sewerageconnections. The policy has usually been to con-centrate primarily on the (subsidized) provision ofwater, often through house connections for thebetter-off and standpipes or handpumps for thepoor.

Consumers in most industrial countries pay allof the recurrent costs (operations, maintenance,and debt service) of both water and sewerage ser-vices. They also pay most of the capital costs of

103

0 50 100 0 50 100 150 200 250 300 350

Quantity conserved (millions of cubic meters per year)

water supply and a large (typically over half) andrising portion of the capital costs of sewerage. Indeveloping countries, by contrast, consumers payfar less. A recent review of World Bank-financedprojects showed that the effective price chargedfor water is only about 35 percent of the averagecost of supplying it. The proportion of total projectfinancing generated by utilities points in the samedirection: internal cash generation accounts foronly 8 percent of project costs in Asia, 9 percent inSub-Saharan Africa, 21 percent in Latin Americaand the Caribbean, and 35 percent in the MiddleEast and North Africa.

A new approach

In urban areas there is abundant evidence thatmost people want on-plot water supplies of rea-sonable reliability and are willing to pay the fullcost of these services. In some areas this standardsolution will have to be adjusted and special ef-forts made to accommodate poor people. In LatinAmerica and, more recently, in Morocco utilitieshave helped poor families to install a connectionand in-house plumbing by giving them the optionof paying over several years. Another option is a"social tariff" whereby the better-off cross-subsi-

Box 5.3 Willingness to pay for water in rural areas

The World Bank, in conjunction with other agencies,recently completed a study of rural water demand inBrazil, Haiti, India, Nigeria, Pakistan, Tanzania, andZimbabwe. The study suggests that where water de-mand is concerned, there are four broad categories ofrural community.

Type I: willingness to pay for private connections is highand willingness to pay for public water points is low. Com-munities in this group offer exciting possibilities be-cause people want and are willing to pay the full costsof reliable water service delivered by way of privatemetered connections into the house or yard. The avail-ability of free public taps (for the poor) will not appre-ciably affect the demand for private connections. Theappropriate strategy is to offer private connections andeven encourage them (specifically, by amortizing con-nection costs in monthly water bills); to recover allcosts through the tariff; and to deliver a reliable ser-vice. A striking finding from the World Bank study isthat this category is larger than is commonly assumed;it probably includes many communities in SoutheastAsia, South Asia, Latin America, and the Middle Eastand North Africa.

Type II: only a minority of households are willing to paythe full costs of private connections, but most households arewilling to pay the full costs of public water points. Althoughoverall willingness to pay for improved water service isconsiderable in Type II communities, users vary greatlyin their willingness to pay for different levels of service.In these villages the provision of free public waterpoints (such as standpipes, wells, or boreholes) wouldsignificantly reduce the demand for private connec-tions. When there is heavy reliance on public waterpoints, some charge must be levied on water fromthese sources in order to finance the system. Here thegreatest challenge is to devise revenue collection sys-tems that are sensitive to peoples' preferences aboutwhen they want to buy water and how they want topay for it. Kiosks appear to be an attractive and flexible

option for many households. Those who wish to havehouse connections should be able to do so but musthave metered connections and must pay the full cost.Many of the better-off communities in Sub-Saharan Af-rica and poorer communities in Asia and Latin Americaprobably fall into this category.

Type Ills households' willingness to pay for improved ser-vice is high but not high enough to pay the full costs of animproved service. This group typically includes poorcommunities in arid areas in South Asia and Sub-Saharan Africa. As in Type II villages, people are will-ing to pay a relatively large share of their income forimproved water service. The distinction is that thecosts of supply are so high, as a result of a combinationof aridity and low population densities, that improvedsystems will not be built and operated without sub-sidies. Given the high priority that people give to im-proved water supply, if transfers were available fromcentral government or from foreign donors, house-holds would typically choose to spend the funds on animproved water supply. The primary service offered insuch communities would be public taps, wells, or bore-holes, although in piped systems metered yard tapsshould be allowed, with tariffs set to recover full costs.

Type IV: willingness to pay for any kind of improved ser-vice is low. This group typically includes poor commu-nities in which (a) traditional water supplies are consid-ered more or less satisfactory by the population or (b)water supply is seen as the financial responsibility ofthe government. In such communities self-financedimproved water supplies are not feasible. Given thelow priority accorded improved water supply, availablesubsidies could be better used in providing other, morehighly valued infrastructural services. For the time be-ing, the appropriate rural water supply policy in suchcases is simply to do nothing. For the second category,once government paternalism ceases, communitiesmay express a willingness to pay and will become TypeII communities.

104

Users want reliable supplies

Figure 5.4 How reliability of supply affectswillingness to pay for piped water:Punjab, Pakistan

Connection rate(percentage of households)100

75

50

25

Source: World Bank Water Demand Research Teamforthcoming.

dize the poor. Properly executed, such policies areboth sensible (since the poor use relatively littlewater) and compassionate. But there are dangers.Social tariffs can lead to a general spread of sub-sidies. And the assignment of noncommercial ob-jectives to a public enterprise generally has an in-sidious effect on the achievement of all itsobjectives, commercial and noncommercial alike.

It is widely assumed that the demand situationin rural areas is quite different, that there peoplehave only a "basic need" which can be met with apublic tap or handpump. But a recent multicoun-

Credit for connections is vital

Figure 5.5 How spreading connection costsover time affects willingness to pay for pipedwater: Kerala, India

Connection rate(percentage of households)100

50

0 5 10 15

Tariff paid by customer(rupees per month)

Revenue collected by supplier(rupees per 100 householdsper month)

1,000

500

20

0 5 10 15

Tariff paid by customer(rupees per month)

Connection cost amortizedConnection cost as lump sum

Source: World Bank Water Demand Research Teamforthcoming.

20

try study by the World Bank of rural water de-mand (Box 5.3) found that most rural people wantand are willing to pay for a relatively high level ofservice (yard taps). As shown in Figure 5.4, theywill pay substantially more if that service is reli-able. And, as Figure 5.5 illustrates, more peoplewill make use of improved water supplies if inno-vative financing mechanisms are employed.

Twenty years of experience with the provision ofwater in rural Thailand (Box 5.4) shows how it ispossible to break out of a "low-level equilibriumtrap" (in which a low level of services is provided,

105

0 20 40 60 80 100

Tariff paid by customer(rupees per month)

Revenue collected by supplier(rupees per 100 householdsper month)

0 20 40 60 80 100

Tariff paid by customer(rupees per month)

Improved reliabilityExisting reliability

Box 5.4 Breaking out of the "low-level equilibrium trap" in northeast Thailand

A well-documented case in northeast Thailand, cover-ing a twenty-year period, demonstrates the importanceof discovering what users of rural water services wantrather than making assumptions about the answers.

Since the people in the area were poor, the initialproject was intended to provide protected water at thelowest possible cost. Because groundwater is abundantin the region, the technology chosen was handpumps.After five years most of the handpumps were notworking, and water use habits were largely un-changed. In a follow-up phase motor pumps providedpiped water at community standpipes. Again, the proj-ect failed. Five years after implementation 50 percent ofthe systems were not working at all, and another 25percent operated intermittently.

As was consistent with conventional assumptions,the failures were attributed to technologies that weretoo complex to maintain and to the inability of the vil-lagers to pay for improved supplies. Gradually, how-ever, it became apparent that the main problem wasnot the capabilities of the villagers but the fact that theservice being offered was not what they wanted. They

did not want handpumps, which were not consideredan improvement over the traditional rope-and-bucketsystem. And standpipes, being no closer than their tra-ditional sources, offered no obvious benefits. Onlypiped water to yardtaps could meet people'saspirations.

In the next project yardtaps were allowed, with theusers paying the full costs of connection. Five yearslater the verdict was in: 90 percent of the systems werefunctioning reliably, 80 percent of the people wereserved by yardtaps, meters had been installed, and lo-cally adapted charging systems had been developed.Not only were the systems well maintained, but be-cause the service was so popular, many systems hadextended distribution lines to previously unservedareas.

In other words, in terms of the typology discussed inBox 5.3, when these (poor) people were treated as"Type IV" cases, the result was the familiar low-levelequilibrium trap. When they were treated as "Type I"communities, the cycle was broken, and a high-levelequilibrium was established.

willingness to pay and thus revenues are low, andthe operation consequently deteriorates) to a

"high-level equilibrium" in which users get a highlevel of service, pay for it, and maintain the de-sired system.

Increasing investments in sanitation

Public investment in water supply and sanitationaccounts for 10 percent of total public investmentin developing countries, or about 0.6 percent ofGDP. Spending on sewerage and sanitation ac-counts for substantially less than one-filth of lend-ing in World Bank-financed projects. Most of thishas been for sewage collection, with little spent ontreatment. An indication of the huge underinvest-ment in treatment is that only 2 percent of sewagein Latin America is treated. Similarly, only a smallproportion (typically 5 percent in developingcountries, compared with 25 percent in industrialcountries) of all spending on solid wastes is di-rected to their safe disposal.

Taking account of demand

There is abundant evidence that urban families arewilling to pay substantial amounts for the removalof excreta and wastewater from their neighbor-

106

hoods. People want privacy, convenience, and sta-tus; polluted water smells unpleasant and fostersmosquitos; and the installation of sewers typicallyincreases property prices. As with water supply,so with sanitation: where public provision is ab-sent, people pay significant amounts for privatelyprovided services. Even in poor cities the amountspaid are considerable. In Kumasi, Ghana, for ex-ample, the use of public latrines and bucket la-trines accounts for large recurrent expendituresabout 2.5 and 1 percent, respectively, of family in-come. In Kumasi and in Ouagadougou families arewilling to pay about 2 percent of household in-come for an improved sanitation system. This isroughly the amount paid for water and for electric-ity. The examples of northeast Brazil and of Or-angi, Pakistan, discussed in Boxes 5.5 and 5.6show the willingness of households to pay for hav-ing wastewater carried out of the neighborhood(by means of a low-cost sewer).

Expanding the menu of supply options

A vital element of a demand-driven sanitationstrategy is to expand the menu of services fromwhich users can choose.

In city centers there is no alternative to costlywaterborne systems. But even in relatively poor

Box 5.5 Innovative sewerage in northeast Brazil: the condominial system

The condominial system is the brainchild of José Carlosde Melo, a socially committed engineer from Recife.The name "condominial" was chosen for two reasons.First, a block of houses was treated like a horizontalapartment buildingor condominiais, in Portuguese.Second, "Condominial" was the title of a popular Bra-zilian soap opera and so was associated with the best inurban life! As is evident in Box figure 5.5, the result is alayout radically different from the conventional sys-tem, with a shorter grid of smaller and shallower"feeder" sewers running through backyards and withthe effects of shallower connections to the mains rip-pling through the system. These innovations cut con-struction costs to between 20 and 30 percent of those ofa conventional system.

The more fundamental and radical innovation, how-ever, is the active involvement of the population inchoosing the level of service and in operating andmaintaining the "feeder" infrastructure. Families canchoose to continue with their current sanitation sys-tem, to connect to a conventional waterborne system(which usually means a holding tank discharging intoan open street drain), or to connect to a "condominial"system.

If a family chooses to connect to a condominial sys-tem, it has to pay a connection charge (financed by thewater company) of, say, X cruzados, and a monthlytariff of Y cruzados. If it wants a conventional connec-tion, it has to pay an initial cost of about 3X cruzadosand a monthly tariff of 3Y cruzados, reflecting thehigher capital and operating costs of the conventionalsystem.

Families are free to continue with their current sys-tem. In most cases, however, those families that ini-tially choose not to connect eventually change theirminds. Either they succumb to heavy pressure fromtheir neighbors, or they find the buildup of wastewaterin and around their houses intolerable once the (con-nected) neighbors fill in the rest of the open drain.

Individual households are responsible for maintain-ing the feeder sewers, with the formal agency tendingonly to the trunk mains. This has several related posi-tive results. First, it increases the communities' senseof responsibility for the system. Second, the misuse ofany portion of the feeder system (by, say, putting solidwastes down the toilet) soon shows up as a blockage in

the neighbor's portion of the sewer. The consequenceis rapid, direct, and informed feedback to the misuser.This virtually eliminates the need to "educate" theusers of the system about dos and don'ts and results infewer blockages than in conventional systems. Andthird, because of the greatly reduced responsibility ofthe utility, operating costs are much lower.

The condominial system is now providing service tohundreds of thousands of urban people in northeastBrazil. The danger is that the clever engineering maybe seen as "the system." Where the community andorganizational aspects have been missing, the technol-ogy has worked poorly (as in Joinville, Santa Catarina)or not at all (as in the Baixada Fluminense in Rio deJaneiro).

Box figure 5.5 Conventional and condominialsewage collection systems

Conventional system Condominial system

a

.

.U

U

U

U

U

U iU

Main sewerStreet sewerHouse sewerBackyard toilet

o Plot

cities the difficulties are not insoluble. In For-taleza, a poor city in northeast Brazil, developersof all high-rise buildings are required to, and do,install package sewage collection and treatmentsystems. The point here is not that this is a goodtechnical solution but that even in a relatively poorcity, developers can easily absorb such costs and

pass them on to those who purchase units in thebuildings.

Beyond the urban core, however, conventionalsewerage systems (with average household costsanywhere from $300 to $1,000) are too expensivefor most developing countries. In recent decadesefforts have been made to develop technological

107

UU

U U

UU

U U

. U

U U

. U

U U

alternatives. Most of this work has concerned theonsite disposal of excreta. Pour-flush latrines andventilated improved pit (VIP) latrines are often thetechnologies of choicethey provide good service(privacy and few odors) at reasonable cost (typ-ically about $100 to $200 per unit), and their instal-lation and functioning does not depend on themunicipality or other organization. At even lowercost, there are yet simpler improvements, such asthe latrine slab program that proved successful inMozambique.

For a variety of reasonshigh housing densities,impermeable soils, and the need to dispose of con-siderable quantities of domestic wastewateron-site solutions do not function well in many urbanareas. Sewage and wastewater collect in the streetsand in low-lying areas, creating serious aestheticand health problems. And in many settings peopleaspire to "the real thing"waterborne sewerage.

Current sanitation choices include a Rolls-Royce(conventional sewerage), a motorcycle (an im-proved latrine), and a bicycle (an unimproved la-trine). What is missing is the Volkswagensome-thing that provides much the same service as theRolls Royce but that many more people can afford.Several such technologies are being developed:

Effluent sewerage is a hybrid between a septictank and a conventional sewerage system. Its dis-tinctive feature is a tank, located between thehouse sewer and the street sewer, that retains thesolids, thereby allowing smaller sewers to be laidat flatter gradients and with fewer manholes. Suchsystems have been widely used in small towns inthe United States and Australia and in Argentina,Brazil, Colombia, India, Mozambique, and Zam-bia. The (limited) cost data suggest that solids-freesewerage costs about 20 percent less than conven-tional sewerage.

Simplified sewerage, developed in São Paulo,allows smaller, shallower, flatter sewers withfewer manholes. This simplified design works aswell as conventional sewerage but costs about 30percent less. It is now routinely used in Brazil.

The condominial system described in Box 5.5has been developed and applied in northeast Bra-zil. It comprises shallow, small-diameter backyardsewers laid at flat gradients and costs about 70percent less than a conventional system.

The Orangi Pilot Project in Karachi (describedin Box 5.6) adapted the principles of effluent sew-erage and simplified sewerage to the realities of ahilly squatter settlement in Karachi. The resultnot just the result of clever engineeringwas adrastic reduction in the cost of sewers, from the

108

$1,000 per household that was standard in Karachito less than $50 per household (excluding the costof the trunk sewers). The achievement is extraordi-naryabout 600,000 people in Orangi are nowserved with self-financed sewers.

Investing in waste disposal

There is an important difference between "privategoods" (including water supply and even waste-water and solid waste collection), in which the pri-mary benefits accrue to individual households,and waste treatment and disposal, in which thebenefits accrue to the community at large. In thefirst case willingness to pay is an appropriateguide to the level of service to be provided, andthe main source of finance should be directcharges to the users. In the case of waste disposal,however, public financing is essential. Govern-ments that subsidize "private" water supply andwastewater collection services are left with lessmoney to finance treatment and disposal services.

No developing country, however, will have theluxury of collecting and treating wastewater fromall households. Because the costs of meeting suchgoals are extremely high, even in industrial coun-tries the full population is not served by wastewa-ter treatment facilities; coverage is only 66 percentin Canada and 52 percent in France. In making theinevitable choices, the best ratio of benefits to costswill usually be achieved by concentrating mostpublic funds on waste treatment in large cities,especially those that lie upstream from largepopulations.

In recent decades some important advanceshave been made in innovative sewage treatmentprocesses. At the lower end of the spectrum is thestabilization pond, a technology that has provedrobust, easy to operate, and (where land is notcostly) relatively inexpensive. A promising inter-mediate (in both cost and operational complexity)is the upflow anaerobic sludge blanket, which hasperformed well in Brazil and Colombia. The pointis the importance of developing technical solutionsthat are adapted to the climatic, economic, andmanagerial realities of developing countries.

Rethinking institutional arrangements

A recent comprehensive review of forty years ofWorld Bank experience in water and sanitationpinpoints "institutional failure" as the most fre-quent and persistent cause of poor performance bypublic utilities. This section deals with the keyareas for institutional reform.

Box 5.6 Innovative sewerage in a Karachi squatter settlement: the Orangi Pilot Project

In the early 1980s Akhter Hameed Khan, a world-renowned community organizer, began working in theslums of Karachi. He asked what problem he couldhelp resolve and was told that "the streets were filledwith excreta and wastewater, making movement diffi-cult and creating enormous health hazards." What didthe people want, and how did they intend to get it?he asked. What they wanted was clear' 'people as-pired to a traditional sewerage system. . . it would bedifficult to get them to finance anything else." Andhow they would get it, too, was clearthey wouldhave Dr. Khan persuade the Karachi Development Au-thority (KDA) to provide it free, as it did (or so the poorperceived) to the richer areas of the city.

Dr. Khan spent months going with representatives ofthe community to petition the KDA to provide the ser-vice. When it was clear that this would never happen,Dr. Khan was ready to work with the community tofind alternatives. (He would later describe this firststep as the most important thing he did in Orangi-liberating, as he put it, the people from the immobiliz-ing myths of government promises.)

With a small amount of core external funding, theOrangi Pilot Project (OPP) was started. It was clearwhat services the people wanted; the task was to re-duce the costs to affordable levels and to develop or-ganizations that could provide and operate the sys-tems. On the technical side, the achievements of theOPP architects and engineers were remarkable and in-novative. Thanks partly to the elimination of corrup-tion and the provision of labor by community mem-bers, the costs (for an in-house sanitary latrine andhouse sewer on the plot and underground sewers in

the lanes and streets) were less than $50 perhousehold.

The related organizational achievements are equallyimpressive. OPP staff members have played a catalyticrole: they explain the benefits of sanitation and thetechnical possibilities to residents, conduct research,and provide technical assistance. The OPP staff neverhandle the community's money. (The total costs of theOPP's operations amounted, even in the project's earlyyears, to less than 15 percent of the amount invested bythe community.) The households' responsibilities in-clude financing their share of the costs, participating inconstruction, and electing a "lane manager" who typ-ically represents about fifteen households. Lane com-mittees, in turn, elect members of neighborhood com-mittees (typically representing about 600 houses),which manage the secondary sewers.

The early successes achieved by the project created a"snowball" effect, in part because of the increasedvalue of properties with sewerage systems. As thepower of the OPP-related organizations increased,they were able to put pressure on the municipality toprovide funds for the construction of trunk sewers.

The Orangi Pilot Project has led to the provision ofsewerage services to more than 600,000 poor people inKarachi and to recent initiatives by several munici-palities in Pakistan to follow the OPP method and,according to OPP leader Arif Hasan, "have govern-ment behave like an NGO." Even in Karachi the mayornow formally accepts the principle of "internal" devel-opment by the residents and "external" development(including trunk sewers and treatment) by themunicipality.

Improving the performance of public utilities

A World Bank review of more than 120 sector proj-ects over twenty-three years concludes that only infour countriesBotswana, Korea, Singapore, andTunisiahave public water and sewerage utilitiesreached acceptable levels of performance. A fewexamples illustrate how serious the situation is:

In Accra only 130 connections were made to asewerage system designed to serve 2,000connections.

In Caracas and Mexico City an estimated 30percent of connections are not registered.

Unaccounted-for water, which amounts to 8percent in Singapore, is 58 percent in Manila andabout 40 percent in most Latin American cities. ForLatin America as a whole, such water losses cost

between $1 billion and $1.5 billion in revenue for-gone every year.

The number of employees per 1,000 waterconnections is between two and three in WesternEurope and about four in a well-run developingcountry utility (Santiago) but between ten andtwenty in most Latin American utilities.

Financial performance is equally poor. A recentreview of Bank projects found that borrowers of-ten broke their financial performance covenants. Acorollary is that the shortfalls have to be metthrough large injections of public money. In Brazil,from the mid-1970s to mid-1980s about $1 billion ayear of public monies was invested in the watersector. The annual federal subsidy to Mexico Cityfor water and sewerage services amounts to morethan $1 billion a year, or 0.6 percent of nationalGDP.

109

Public utilities play a dominant role in the provi-sion of water and sanitation services throughoutthe world. There are many examples of such util-ities working effectively in industrial countriesand, as described above, a few cases in developingcountries. An essential requirement for effectiveperformance is that both the utility and the regula-tory body (essential for such natural monopolies)be free from undue political interference. In thecase of the utility the vital issue is managerial au-tonomy, particularly as regards personnel policies;in the case of the regulatory body, it is the settingof reasonable tariffs. Although this recipe is simpleand has been well tested in many industrial coun-tries, it has been extraordinarily difficult to imple-ment in developing countries other than thosewith high levels of governance. Sometimes util-ities and regulators are nominally autonomous,but usually key policies (on investments, person-nel policies, and tariffs, for instance) are effectivelymade by government and heavily influenced byshort-term political considerations.

Many projects financed by external agencieshave addressed the problems of public water util-ities through sizable action plans, technical assis-tance components, and conditionality. Some ofthese efforts, such as that undertaken recently bySri Lanka's National Water Supply and DrainageBoard, have led to significant improvements inperformance. As with public enterprises in othersectors, however, most of these efforts failed be-causein the words of a recent Bank review"public enterprises . . . are key elements of pa-tronage systems, . . . overstaffing is often rife, andappointments to senior management positions arefrequently made on the basis of political connec-tions rather than merit." And things have beengetting worse rather than better. Achievement ofinstitutional objectives in World Bank-financedwater and sanitation projects fell from about twoin three projects in the late 1970s to less than onein two projects ten years later.

Improving the performance of public utilitiesnevertheless remains an important goal, for tworeasons. First, in the medium term public utilitieswill continue to provide services to many. Second,improvement in the performance of public utilitiesis often a precondition if private operators are to beinduced to participate.

Separating provision and regulation

Experience in industrial countries shows that acentral problem in improving environmental qual-ity is that the public sector acts both as supplier of

110

water and wastewater services and as environ-mental regulatorit is both gamekeeper andpoacher. The results of this conflict of interest aresimilar throughout the world. In England andWales prosecutions of those responsible for sew-age treatment were rare when the river basin au-thorities were responsible for water resource man-agement, environmental protection, and services.In 1989 private companies were given responsibil-ity for the delivery of water and sewerage services(with public agencies retaining regulatory author-ity). Since then, fines have been increased sub-stantially and violators have been prosecuted. Theother side of the separation of powers is that ser-vice delivery agencies are, in the process, liberatedfrom serving multiple tasks and can pursue well-defined and specific objectives.

Expanding the role of the private sector

Increased private sector involvement is warrantedin two areas. One is in services to public utilities.In industrial countries the engineering of publicworks is dominated by private firms, which de-pend for their survival on their reputation for per-formance and which assume legal liability for theconsequences of any professional negligence.These factors provide powerful incentives for sup-plying cost-effective, high-quality services andconcurrently furnish a stringent environment forthe supervised apprenticeship training that is a re-quired part of professional certification in thesecountries. By contrast, in many developing coun-tries (particularly in Asia and Africa) the engineer-ing of public works is dominated by large publicsector bureaucracies. Employment security is total,promotion is by seniority alone, good work goesunrecognized, poor work is not subject to sanc-tions, and an atmosphere of lethargy prevails. Thedirect consequence is the construction of high-cost, low-quality facilities; the indirect effects in-clude a weak professional labor force. The obviousanswers are, first, to decrease the direct involve-ment of the government in public works and, sec-ond, to nurture a competitive engineering consul-tancy sector.

More private involvement in the operation ofwater, sewerage, and solid waste companies isalso warranted. Many industrial countries havefound it difficult to reform public enterprises, ex-cept as part of a move to privatize them. Indeed,privatization is increasingly seen as a way not onlyto effect performance improvements but also tolock in the gains.

In developing countries there has been some ex-

perience with private sector operation of waterand sanitation utilities. Côte d'Ivoire has been apioneerSODECI, in Abidjan, is considered oneof the best-run utilities in Africa. After Macao'swater utility was privatized in 1985, performanceimproved dramatically; the percentage of unac-counted-for water fell by 50 percent over six years.Guinea, which recently let a lease contract for sup-plying water to its principal cities, experienceddramatic improvements in the financial conditionof the utility in just the first eighteen months as aresult of raising the efficiency of bill collection from15 to 70 percent.

Other countries have taken more incrementalapproaches. EMOS, the utility serving Santiago,has used private contracts for functions such asmeter reading, pipe maintenance, billing, and ve-hicle leasing. As a result, it has a high staff produc-tivity ratethree to six times higher than that forother companies in the region. Many other coun-tries, faced with persistently poor performance bytheir public utilities, are seriously consideringgreater private sector involvement, following, ingeneral, variations of the French model. For exam-ple, in Latin America, concession contracts are cur-rently being let for the supply of water and sew-erage services in Buenos Aires and Caracas.

Private involvement in the sector is not a pan-acea and is never simple. In the United Kingdomwater privatization is generally considered themost complex of all privatizations undertaken. Indeveloping countries there are formidable prob-lems. For the private operator the risk involved istypically high. In addition to the obvious politicaland macroeconomic risks, knowledge about thecondition of the assets is usually only rudimen-tary, and there is uncertainty about the govern-ment's compliance with the terms of the contract.Groups such as existing agencies and labor unionsthat stand to lose from greater private sector in-volvement often strongly oppose privatization.

For the government, too, there are problems.Because of economies of scale, it is virtually impos-sible to have direct competition among suppliersin a specific area. Countries have tried a variety ofsolutions: in France, there is periodic competitionfor markets, and in England and Wales, economicregulators reward efficiency by comparing the rel-ative performance of different companies (a prac-tice that is unlikely to be applicable elsewhere). Inaddition, in many developing countries it is oftendifficult to attract private sector interest. Only ahandful of firms compete internationally for suchcontracts.

The case for private sector involvement is stron-ger still in the solid waste collection business.Whereas foreign control of water supply is oftenperceived to involve losing sovereignty over a stra-tegic sector, no one cares if foreigners pick up thegarbage. In addition, for populations of more thanabout 50,000 there are no economies of scale andthus no natural monopoly. Experience in manycountriesincluding Argentina, Brazil, Canada,Chile, Colombia, Japan, Switzerland, and theUnited Stateshas shown that the private sectoralmost invariably collects solid wastes more effi-ciently than municipalities. Unit costs for publicsystems are 50 to 200 percent higher, with the pri-vate sector efficiency gains apparently greatest inthe developing countries listed.

Increasing community involvement

Community groups and other NGOs also have animportant role to play in the supply of water andsanitation services and the collection of wastes. Asthe condominial (Box 5.5) and Orangi (Box 5.6)examples show, in the urban fringe the most pro-ductive relationship between community groupsand the formal sector is that of partnership, withthe formal sector responsible for the "external" or"trunk" infrastructure and the community payingfor, providing, and managing the "internal" or"feeder" infrastructure.

Because many water and sanitation services aremonopolies, consumers cannot force suppliers tobe accountable by giving their business to a com-petitor. To give consumers a voice in the politicalprocess, consumers' associations and ratepayers'boards are vital. Paradoxically, because there issuch an obvious need for oversight of the activitiesof a private operator of a natural monopoly,greater private sector involvement stimulatesgreater consumer involvement. In the UnitedKingdom, for example, water users have had amuch greater say in running the industry sinceprivatization.

In recent years external agencies and govern-ments alike have become aware that in rural areasinvolvement of the users is essential if water sup-plies are to be sustained. Generally it has been as-sumed that support to rural communitiesin theform of information, motivation, and technical as-sistancewill come from the government. The diffi-culty is that governments, especially in rural areas,are often weak, and their officials rarely have anincentive to provide support. Here the private sec-tor (including NGOs) may be able to help.

111

Good policies can make a big difference in quality of life

Figure 5.6 Safe water and adequate sanitation: three scenarios, 1990-2030

Population without safe water Population without adequate sanitation

Billions Billions

3.0

2.5

2.0

1.5

1.0

0.5

01990 2000 2010 2020 2030

3.5

3.0

2.5

2.0

1.5

1.0

1990

"Business as usual" scenarioScenario with accelerated investment in water supply and sanitation services a

Scenario with accelerated investment and efficiency reformsb

2000 2010 2020 2030

Note: Assumptions are as follows: growth of per capita income and population as in Chapter 1; per capita income elasticity, 0.3; price elasticity, -0.25;initial prices 60 percent of marginal costs, gradually rising to efficiency levels over a twenty-five-year period; initial supply costs 50 percent higherthan with good practices (due to managerial inefficiencies), gradually being reduced in step with price efficiency reforms; and marginal costs risingat 3 percent per year.

Investment in water supply increases 30 percent, and investment in sanitation services increases 50 percent over the period.To realize this scenario in low-income countries, efficiency reforms - and the resulting increase in investment shares - would need to be greater

than average.Source: World Bank estimates, based on Anderson and Cavendish, background paper.

Several promising examples of the involvementof small-scale private operators in developingcountries have emerged:

In rural Pakistan about 3 million families havewells fitted with pumps, many of which are mo-torized. The water supplies are paid for in full bythe families, and all the equipment is provided andserviced by a vibrant local private sector industry.

In Lesotho the government trained brick-layers to build improved pit latrines. Governmentbanks also provided (unsubsidized) credit to fi-nance the latrines. The program has been a singu-lar success, thanks mainly to the aggressive role ofthe bricklayers in expanding their markets (andproviding services as well).

In West Africa a private handpump manufac-turer has developed a "Sears Roebuck"-typescheme whereby purchase of a pump comes withfive years of support, including training and the

112

provision of spare parts. Later on, the communitywill be able to maintain the pump and will pur-chase the necessary spare parts from local traders.Because the private sector agent has clear incen-tives for providing services effectively, this ar-rangement may work better than government sup-port to the communities.

Finally, women have a central role to play inthese reforms. In most countries the collection ofwater has been considered "women's work" (ex-cept where the water is sold!). Only recently, how-ever, have systematic efforts been made to involvewomen in project identification, development,maintenance, and upkeep. The results have gener-ally been encouraging. In an urban slum in Zambiaa women's organization improved drainagearound public taps. Women have been trained ascaretakers for handpumps in Bangladesh, India,Kenya, Lesotho, and Sudan. In Mozambique

women engineers and pump mechanics performalongside, and as effectively as, their male coun-terparts. In Sri Lanka women's cooperatives havebeen set up to assemble and maintain a locallymanufactured handpump. Women's cooperativesmanage communal standpipes and collect moneyto pay for metered supplies in Honduras, Kenya,and the Philippines. Women who are trained tomanage and maintain community water systemsoften perform better than men because they areless likely to migrate, more accustomed to volun-tary work, and better trusted to administer fundshonestly.

Creating an enabling environment

This chapter has argued that massive improve-ments can be made in health, economic efficiency,and equity through better provision of sanitationand water. The key is firmly in the hands of gov-ernments, for the single most important factorneeded is political will. Where there are long-es-tablished and deeply entrenched traditions ofsound governance (as in Botswana, Korea, andSingapore), it is evident that autonomous, ac-countable public sector agencies can provide effi-cient and equitable service. For many countries,however, such levels of governance are not attain-able in the short run, so that greater involvementof the private sector and NGOs will be crucial tothe provision of accountable and efficient services.

To allow helpful change to occur, the govern-ment must concentrate on the things that it, andonly it, can do. Its job is to define and enforce anappropriate legal, regulatory, and administrativeframework. This includes tasks as fundamentaland diverse as rewriting legislation so that water

markets can come into existence, rewriting con-tract laws so that the private sector can participatewith confidence, building a capacity for environ-mental and, where appropriate, economic regula-tion, developing financial mandates for utilitiesthat encourage conservation, and setting and en-forcing quality standards for equipment. The gov-ernment must also create conditions under whichothersthe private sector, NGOs, communities,and consumerscan play their parts.

What might be accomplished

More than 1 billion people are still without accessto safe water and 1.7 billion people are withoutaccess to adequate sanitation facilities. Elementarycalculations show that an "unchanged practices"or "business-as-usual" scenario would lead to arise in the number of people without service in thecoming decades (the top curves in Figure 5.6). Thisis a result of rising unit costs, as well as unprece-dented increases in population. If the shares oftotal investment allocated to sanitation (currently0.6 percent of gross investment) and to water sup-ply (currently 1.7 percent) were raised by, say, 50and 30 percent, respectively, the numbers un-served might still rise, although not as much (themiddle curves in the figure). Far more important(as shown by the bottom curves) is the combina-tion of policy reforms and accelerated investment.By attracting financial, managerial, and skilled la-bor into the sector and by freeing enterprises toinvest more and improve maintenance, this newapproach, which is already being adopted in somecountries, could bring about dramatic increases inaccess to sanitation and clean water within thenext generation.

113