Cities of the developing world face a formidableundertaking, given the rapid rate of growth andsheer numbers of urban residents to be em-

ployed, housed, and serviced. Cities offer proximity,which generate externalities both positive and negative.On the positive side, proximity is a source of produc-tivity; industrial and service activities emerge in citiesbecause entrepreneurs and small firms can share mar-kets, infrastructure, labor, and information. Cost sav-ings and productivity advantages that accrue to firmswhen they locate near each other in the same industryor near other economic activities derive directly fromphysical proximity or indirectly from less tangible in-teraction among economic actors (learning and net-working, leading to innovation). However, large groupsof people and activities in close proximity also generatenegative externalities: poor sanitation, pollution of airand water, congestion, crime, and so on. This puts apremium on the quality of institutions—both formaland informal—to ensure the positive externalities andto cope with the negative externalities.

As the previous two chapters have noted, the de-velopment of urban areas will have to be better co-ordinated with the development of rural areas byproviding markets for rural products, by subcon-tracting activities to expand nonfarm rural employ-ment, and by helping rural migrants adapt morerapidly to city life. Cities and towns facilitate soci-ety’s transformations in knowledge, institutions, andeconomic activity. By bringing together diverse peo-ple and activities, urban areas offer great opportuni-ties for improving the quality of life.1 For cities andtowns to realize the promises of a better life—espe-cially for poor people and for migrants from rural

C H A P T E R 6

Getting the Bestfrom Cities

areas—they need stronger institutions to providewide access to assets and to balance interests that en-sure the provision of public goods. Such institutionsare central to an urban governance that is inclusiveof all residents, responsive to their needs, and con-ducive to careful management of natural resourcesand wastes.

This chapter first describes the opportunities andchallenges for urban life and then asks the followingquestions:

� How can informed constituencies be built to ad-dress spillovers and anticipate risks? Providing in-formation, building knowledge, and mobilizingdispersed interests are key to creating constituentsthat act together to anticipate problems and toprevent and manage disasters.

� How can competing interests be balanced anddispersed interests articulated to provide urbanpublic goods? Foresight, political will, and a gov-ernance system that is accountable to a wide arrayof stakeholders are key ingredients for achievingcredible commitments.

� How can inclusion and access to assets be encour-aged—one key to a city’s sustainability? Securityof tenure and guidance of new settlements toprevent future slums will lessen the inequitableaccess to assets, thus empowering and enablingpoor people to become productive members ofurban society.

� What institutional mechanisms are necessary forgood urban governance and sustainable urban de-velopment? These include an appropriate sharingof responsibility and coordination across stake-

holder groups; wide participation in strategic plan-ning; and networking among practitioners andstakeholders.

City lights: beacons of hope and warning flares

The rising share of people in urban areas and thecorresponding economic growth of cities and townsare two defining experiences of economic and socialdevelopment (box 6.1). Urban areas offer possibili-ties for greater welfare because they give individualsthe opportunity (through a myriad of functioningurban markets) to develop a wider and larger port-folio of assets—and to achieve higher returns to theirlabor. They also exist because of collective concernsto share culture, learning, religious observation, andmutual protection.

Yet across cities in all regions of the world there isevidence that the potential benefits, both individualand collective, are not being fully realized, and areclouded instead by myriad problems. The inadequateprovision of jobs, housing, and other goods and ser-vices stems from imperfections in markets and poli-cies. For many environmental and social concerns,markets cannot provide the coordination needed toreveal interests and minimize transaction costs. Inmany countries institutional failures mean that mar-kets are less effective than they could be, while alter-native solutions and innovative uses of market in-struments to address the threats to sustainabledevelopment are inadequately developed. This chap-ter does not examine the full range of good urbanpolicies.2 Instead, it focuses on the conditions forbuilding institutions to protect urban assets—partic-ularly environmental and social assets—becausethese conditions shed light on the potential for iden-tifying and adopting good policies, and determiningsociety’s ability to respond to future concerns.

The role of cities in sustainable developmentOver the next three decades the urban population ofdeveloping countries will grow (from natural in-crease, migration, and reclassification of formerlyrural areas) by 60 million people a year—equivalentto the population of the Arab Republic of Egypt or Ethiopia. Urban areas will need to perform a few key functions to support sustainable nationaldevelopment:

� Facilitate social and institutional change, by im-proving access to ideas, knowledge, and technol-

ogy and by creating both the impetus and oppor-tunity for innovation.

� Provide employment and services at a scale suffi-cient for current residents and new arrivals. Pro-ductive employment is critical so that the fall indemographic dependency ratios projected in mostdeveloping countries over the next 20–30 years cantranslate into increased savings and investment.

Two categories of urban issues are relevant to this Report:the spatial system of urban areas in the country and theperformance of urban areas. The first topic is discussed inchapter 7. This chapter takes the second category of urbanissues—the city and its governance—as its unit of analy-sis. Focusing on the city spotlights the impacts of urban liv-ing (good and bad) and how the relevant institutional frame-work, both national and local, affects these outcomes.

Although the analysis of this chapter applies both tocities and to towns, as their smaller counterparts, the dis-cussion mainly refers to cities. Most of the benefits of ag-glomeration and most of the diseconomies appear astowns become recognizable cities.

The Report examines changing opportunities and chal-lenges that appear as population moves along the set-tlement-size continuum, but it does not argue that theprospects for sustainability are a function of scale. Verylarge settlements (cities of multimillion inhabitants) areneither necessarily the best nor the worst cases of sus-tainable development.

� In some countries urban air quality tends to be worst in medium-size cities (populations between 100,000–500,000).*

� Crime and vulnerability to disaster often become prob-lems in urban agglomerations well below 1 millioninhabitants, but do not increase proportionally with pop-ulation size.

� Congestion tends to worsen with city size but also de-pends on such other factors as public transport, trafficmanagement, and road space.

� Many of the economic benefits of urban productivity,such as higher wages and increased human capital, ap-pear positively correlated with city population, at leastto a fairly large threshold.

What is most clear is that the quality of urban gover-nance and management is critical to gaining the benefitsand reducing the negative aspects of cities of any size. Aschapter 1 notes, the projected trend of increasing num-bers of people (and possibly shares of urban population)living in very large cities over the coming decades in de-veloping countries will put a premium on building institu-tions to address the problems of those cities.

* Observation based on countries with most extensive data bycities. Lvovsky (2001), annex B.

Box 6.1

The focus of “urban” in this chapter

� Ensure a healthful and attractive environment forthe urban population while protecting natural re-sources and reducing deleterious impacts on widerregions and later generations. The massive new in-vestment in the capital stock of cities required forthe doubling of urban population by 2030 will becritical to environmental outcomes.

What enables urban areas to promote change andimproved quality of life is their scale and density—and their social and economic diversity. Proximityand heterogeneity make urban areas mechanisms forknowledge and learning, for productivity and mar-ket development, and for improved choice and qual-ity of services. And they do this even more when theinstitutional conditions are right.

The urban stimulus to social transformation andinnovation. The shift from rural to urban society,with greater mingling among diverse people, trans-forms social attitudes and behaviors. It reveals thelimits of traditional values and institutions and in-tensifies pressures for change in local governance andintergovernmental relations. Traditional social normsthat perpetuate inequalities for women and for cer-tain minorities tend to be less strictly enforced in theurban environment.3 Urban households are generallymore motivated to limit family size, because of eco-nomic alternatives and lifestyles.

The ferment of urban life generates new forms ofcollective action to address the challenges that arise.Urban-based constituencies have been the drivingforce behind many of the environmental causes thatpertain to national and global public goods—creat-ing national parks, protecting biodiversity, and man-aging coastal zones. More than 90 percent of China’senvironmental NGOs are located in cities of the rel-atively well-developed eastern coastal region.4 A re-cent study of Indonesia proposed that the best wayto save the tigers was to teach urban children aboutthem.5

Historically, cities have been centers of learningand innovation. The growing intensity of knowledgeexchanges arising from globalization and the infor-mation technology revolution has the greatest im-pact where there is also occasion for interpersonalcommunications.6 Informal information or tacitknowledge, important to productivity and to socialrelationships, thrives on face-to-face contacts.7 Re-search among Mexican firms, for example, showsthat access to informal networks (business lunches

with local buyers, suppliers, competitors, and gov-ernment officials) has a significant and positive ef-fect on their productivity.8 And proximity to highereducation institutions provides firms with opportu-nities to commercialize research ideas, often throughuniversity-enterprise partnerships.9

Sources of urban productivity. Urban employmentand services benefit from the economies of agglom-eration—from cost savings and other advantages thataccrue to firms when they locate near others in thesame industry, or simply near other economic activi-ties to share markets, services, infrastructure, labor,and information. The productivity advantage meansthat urban investment has strong multiplier effectsin stimulating other high-value activities. The bene-fits extend to rural areas, which need access to urbanmarkets to expand and diversify both farm and non-farm production.

As a rule, larger urban areas are the most produc-tive since they allow for greater specialization inlabor use, better matching of skills and jobs, and awider array of consumption choices for workers andancillary services for producers.10 As long as thisgreater productivity outweighs higher costs for land,labor, housing, and other necessities, the city canthrive. Once the diseconomies become too great,larger cities may lose their edge in creating jobs orimproving the welfare of residents, unless they canshed some activities (those that are more mature andstandardized) to smaller cities to make room for oth-ers (more innovative and higher value industry andservices) and change land uses.

For cities to fulfill their potential as engines of na-tional economic growth, they need to ensure that thelabor market is not only deep but well integrated andinclusive—with accessible workplaces and residences.A city can improve its investment climate. However,cities in general can only improve the national invest-ment climate if their overall legal and regulatoryframework complements the national framework tominimize risks, uncertainties, and transaction costs toinvestors. This is especially important for small andinformal sector enterprises, which provide most urbanemployment, rely more heavily on publicly providedinfrastructure and information, and are particularlyvulnerable to institutional and policy failures.

More affordable and higher quality services. Thegreater scope for competition and specialization inall goods and services enables urban areas to provideconsumer benefits in the form of greater choice and

quality.11 But the advantage is especially importantfor services with high fixed costs (increasing returnsto scale), such as middle- and higher-level educationand health facilities, and network infrastructure.12

The cost advantage explains some of the manifestlybetter social indicators in urban areas and in coun-tries that are more urbanized.13 That is, urbanizationhas a large positive impact on a country’s efficiencyin achieving health and education outcomes.14 Evenwhere more expensive services are required for envi-ronmental and health benefits (as for waste disposal,sewerage, water treatment, and mass transit), theadded cost can often be justified by higher economicreturns.

Migrating for choice and change. The offer of newopportunity and a better life is often what draws mi-grants to towns and cities (the pull factor). Limitedemployment options in rural areas, whether fromagricultural dislocations arising from natural and so-cial disasters, or from increased agricultural produc-tivity that reduces the demand for farm labor (chap-ters 4 and 5), also contribute (the push factor).15

Cities and towns allow individuals to substitute theirhuman capital (work effort and skills) for natural, fi-nancial, or physical assets they may lack—and to

more steadily transform and expand their portfolioof assets than is possible in many rural areas.

Those who migrate to cities are often better offthan their neighbors back home;16 in cities they re-ceive more education and better skills, and in thelonger run they catch up with established urban res-idents.17 A recent survey in Latin America finds apotentially large private gain from migration tourban areas, in part because the returns to humancapital tend to be larger there.18 A study of the urbanlabor market in the Punjab State of India found noevidence that migrants remain confined to marginaljobs or are disproportionately unemployed.19 Mi-grants frequently obtain work, housing, and urbanservices through the informal sector, and they oftendepend on supportive social networks to do so (box6.2). Nonetheless, imbalances in the demand andsupply of jobs, housing, and urban services can con-tribute to urban poverty initially and for long periods.

In the short term, the influx of migrants cansometimes overwhelm some urban areas, particularlywhen the pace is sharply accelerated by civil conflictor national disaster. In the medium term, as part ofthe natural development process, rural-to-urban mi-gration and the resulting return flow of transfers

Urban areas are often said to lack social capital. Yet social net-works are important to the survival and mobility strategies ofurban poor people and to decisions about migration from ruralareas. Risks in the urban context arise mainly from weak prop-erty rights (which can result in loss of assets and involuntaryresettlement), inadequate sanitation, exposure to violent crime,and unemployment or other effects of macroeconomic shocks.The urban poor, like their rural counterparts, cope by using theirsocial networks and personal assets.

Social networks in the city are based more on reciprocallinks between individuals and friends, than on familial obliga-tions (as in rural areas). Yet, maintaining close links betweenrural and urban social networks can be crucial to preservingone’s identity. As in rural areas the ability of urban commu-nities to engage in collective action is often instrumental in obtaining public services. This is particularly the case in largecities (for instance, Jakarta or Manila) where the urban poor—because of their numbers, and the relative ease of organizingthem—are an important political constituency that can also bemanipulated by officials promising services.

Recent field research in the slums of Delhi confirms that themajor source of risk to the home and asset base for residentsof squatter settlements is insecure land tenure. The slums with

the least security are usually those harboring new migrants andothers with relatively limited social networks, who are unableto negotiate with bureaucrats and politicians. Social networksamong slum residents assist with survival (coping with emer-gency needs), similar to traditional rural networks, and to up-ward mobility of individuals and the community. The transitionfrom an “unrecognized” to a “recognized” slum affords accessto water and sanitation and immunity from demolition. Similarly,networks are used to get ahead occupationally, whether throughformal sector employment or informal self-employment. Slum-dwellers also use relationships with local leaders to obtain citi-zenship status and strengthen their legal protection.

The urban poor, including migrants, move from their initialinherited networks to ones that link them to external benefitsand resources outside their original community. While thesenetworks are a useful resource, the energies of poor peoplewould be used more productively if basic services and securityof tenure were provided more consistently and transparently,thus removing both the need for exhaustive negotiations to ac-quire basic entitlements and the opportunity for rent-seekingby local officials.

Source: Rao and Woolcock (2002).

Box 6.2

How social networks help the urban poor manage risks and get ahead

raise incomes and living standards in both areas. Cir-cular and temporary migration to and from cities ortowns help manage risk for both rural and urbanhouseholds.20 Openness to new ideas and learningacquired by urban migrants are transmitted to ruralcommunities through social and family links—andthrough the use of remittances to introduce technol-ogy to rural activities.21 In the longer term, once theurban transition has been completed, natural popu-lation increase in cities, rather than rural-to-urbanmigration, will account for most urban growth. Mi-gration among cities will continue in response tochanging economic opportunities.

Making the environment work for urban resi-dents—and saving it for others. Urban living poses en-vironmental hazards, which affect the current popu-lation (especially poor people) through immediate,local impacts on health and safety. It also causes en-vironmental degradation, with longer-term, wider-area, and intergenerational consequences.22 Varia-tions in the incidence and relative severity of a rangeof environmental problems across cities at differentlevels of development suggest differences in priori-ties for action (table 6.1).

In low-income cities fewer than half the house-holds are connected to water and sewerage, and percapita water consumption is half that of cities withlower-middle income ranking (table 6.2). Less thanone-third of solid waste in the poorest cities is dis-posed of properly; only the richest cities providecomprehensive wastewater treatment. Partly reflect-ing environmental risks, the average mortality ofchildren under five in the poorest cities is more thantwice that in the next city-income category, and 20times that in the richest cities.

Especially in cities at low levels of developmentmany residents face environmental risk because oftheir living conditions and location. These house-holds are least able to afford protective or mitigatingmechanisms—or to assert claims for improved ser-vices. Most vulnerable are children and women, theelderly and disabled, and homeworkers, who arecontinuously exposed to hazards in their immediateenvironment.

In the poorer cities badly managed urban growthdegrades natural resources, especially watersheds,soils, and coastal environments—because of un-treated sewage discharge, poor solid waste disposal,and a lack of storm drainage. In contrast, many of

the issues of environmental degradation in richercities, such as greenhouse gas emissions from hydro-carbon fuels, stem from lifestyles entailing high con-sumption and associated waste of natural resources.

Despite these differences in the incidence of risksand their links to income and consumption, urbanresidents of middle-income countries suffer environ-mental insults both traditional and modern (such asexposure to hazardous wastes and chemical pollu-tants).23 International travel and changing globalweather patterns are bringing environmental profilesof cities in industrial and developing countries closertogether by spreading some risks (of disease and natu-ral disasters, respectively) to both groups of countries.For the full range of concerns, institutional reform is required to protect poor people and environmentalassets in cities both today—and in the future.24

Allowing the urban potential to transform societyand to improve welfare, while also protecting the en-vironment, may appear to be harder for the develop-ing world today than it was for industrial countriesat similar points in their urban transition. Why? Be-cause of today’s faster urban population growth rates(figure 6.1) (approximately one-half from migrationand the rest from natural growth and reclassificationof contiguous areas) and the sheer numbers of urbanresidents to be employed, housed, and serviced overthe next few decades. Urban population growthrates, especially in Sub-Saharan Africa, have beenunprecedented, though as in other regions they areprojected to slow. Despite the deceleration, almost20 million new urban residents a year are projectedthrough 2030 for East Asia.

The same characteristics of urban areas—density,scale of settlement, and social diversity—that canbring about the positive potential of more jobs, ser-vices, and learning also create the negative potentialand the need to balance interests. In addition to en-vironmental spillovers, urban areas are associatedwith other threats to sustainable development andlivability. Problems of land use and accessibility (in-cluding congestion) impede the utility, inclusiveness,and enjoyment of urban life, while physical insecu-rity results from poorly managed risks of natural dis-asters, crime, and violence. All of these problemsaffect poor people even as they reduce the welfare ofall urban residents.

None of these threats is driven primarily by therate of urbanization, by the size of city, or by the lack

Table 6.1

Urban environmental issues and status by level of city development

Sector or problem area Low Lower-middle Upper-middle High

Note: Cities grouped by estimated city product (city average income calculated by national accounts methods). Sample is of cities (including inOECD countries) with available data and is not statistically representative. Low income defined as city product below $750 per capita a year;lower-middle as $751–2,499; upper-middle as $2,500–9,999; high as above $10,000.Source: Adapted from Leitmann (2001) and Hardoy, Mitlin, and Satterthwaite (2001).

Water supplyservice

Sanitation

Drainage

Water resources

Solid wastemanagement

Air pollution

Greenhouse gasemissions

Land management(environmentalzoning of fragilesites and prepara-tion for newsettlements)

Accident risk

Disastermanagement

Low coverage, highbacteria contamination,inadequate quantity forhygiene (high risk of food contamination andinfectious diseases)

Very low coverage, opendefecation in some neigh-borhoods and low ratiopublic toilets to residents; high risk of diarrhealdiseases

Storm drains very in-adequate, poorly main-tained; frequent flooding,creating high risk ofwater-related diseasevectors (mosquitos)

Mixed sewerage andstorm water runoff towater bodies causing bac-terial pollution and silting

Little organized collection;recycling by informal sec-tor, open dumping orburning of mixed wastes;high exposure to diseasevectors (rats, flies)

Indoor and ambient airpollution from low-qualityfuels for household usesand power generation

Very low per capita

Uncontrolled land devel-opment; intense pressurefrom squatter settlementson open sites

In-home and workplaceaccidents due tocrowding, fires

Natural disasters producemassive loss of life andproperty especially in set-tlements in disaster-proneareas; little capacity formitigation or emergencyresponse

Low access by poorresidents and informalneighborhoods

Better coverage oflatrines and public toilets,but poorly maintained;low sewerage coverage

Somewhat better than in low income

Risk of groundwatercontamination from poorly maintained latrinesand untreated sewage

Moderate coverage ofcollection service, littleseparation of hazardouswaste; mostly uncon-trolled landfills

Growing ambient air pollu-tion from industrial andvehicular emissions (highper-vehicle, due to ineffi-cient fuels and vehicles)

Low but growing percapita

Ineffective or inappropri-ate land-use controls,pushing new settlementstoward urban periphery;continued high populationgrowth

Increased risks of indus-trial workplace and trafficaccidents (pedestriansand nonmotorizcdvehicles)

Somewhat better than in low-income, althoughwith increasing risk ofindustrial disasters

Generally reliable, butrising demand causingshortages in resourcesupply

More access to improvedsanitation, but still largenumbers of residents inlarge cities not coveredespecially in informalsettlements; most waste-water discharge untreated

Better drainage;occasional flooding

Private wells drawingdown groundwater;severe pollution fromindustrial and municipaldischarge

Better organized collec-tion; severe problems but growing capacity forhazardous waste manage-ment; semicontrolledlandfills

Ambient air pollution stillserious (but greatercapacity to control espe-cially industrial sources)

Rapidly increasing, mainlydue to motorization

Some environmentalzoning

Transport accidents in-creasing, but some miti-gation and emergencytreatment response

Increasing awareness and capacity for disastermitigation and emergencyresponse

Good supply but high totalconsumption; some con-cern with trace pollutants

Full coverage; mostwastewater treated

Good drainage; verylimited flooding

High levels of effluentcontrols and treatment to reduce pollution

Increased emphasis ontotal waste reduction,resource recovery, andpreventing hazardouswaste; controlled landfillsor incineration

Ambient air pollutionmainly from vehicles (due to high volume ofvehicle kilometers)

Very high per capita

Regular use of environ-mental zoning; little popu-lation growth, but risingincomes press for moreland consumption forexisting residents

Rate of industrial andtransport accidents re-duced despite increasingtravel (vehicle kilometers)

Good capacity formitigation and response

Table 6.2

Environmental health, welfare, and living conditions vary by city product

City product category

Lower- Upper-

Indicator Low middle middle High

Household connections (percent)water 48.0 78.8 92.9 99.9sewerage 45.6 68.7 84.3 99.9electricity 72.3 93.6 95.0 100.0

Water consumptionliters per person per day in all settlements 88 161 232 247

Wastewater treatedpercent treated 29.4 56.7 68.2 97.4

Solid waste disposal (percent)sanitary landfill or incinerated 30.7 41.4 37.6 77.7other (open dump, recycled, burned) 65.9 58.3 62.2 22.3

Under-five mortality per 1,000 104.2 39.7 25.8 5.2Households below the locally defined poverty line (percent) 31.7 23.2 16.0 6.9Sample size (cities) 49 36 25 20Note: Cities grouped by estimated city product (city average income calculated by national accounts methods). Sample is of cities (including inOECD countries) with available data and is not statistically representative. Low income defined as city product below $750 per capita a year;lower-middle as $751–2,499; upper-middle as $2,500–9,999; high as above $10,000.Source: United Nations–Habitat Global Urban Indicators Database 1998.

0

0 10 20 30 40 50 60 70 80 90

1

2

3

4

5

6

National urban population share (percent)

Average urban growth rate (percent)

Sub-Saharan Africa

All developing countries

All developing

countries

Developed countries

(1900–25)

South Asia

South Asia

Europe andCentral Asia

Europe andCentral Asia

East Asiaand Pacific

East Asia and Pacific

Middle Eastand North Africa

Middle Eastand North Africa

Sub-Saharan Africa

1975–20002000–25

Latin Americaand the Caribbean

Latin Americaand the Caribbean

Figure 6.1

Many developing countries are undergoing urban transition with relatively high urban population

growth rates

Note: All averages weighted by population. Lines indicate increase in share of urban population between end-point years (25-year increments).Source: Developed country data from Brockerhoff and Brennan (1998); other data and projections from U.N. (1999).

of fiscal or other resources—though these factors(and such others as geography, local culture, and his-torical legacy) can make problems harder to manage.Cities need not suffer crippling diseconomies what-ever their size or income (see box 6.1). Shanghai, oneof the world’s largest urban areas in one of the poor-est countries, devotes more than 3 percent of the cityGDP to environmental protection. And it has man-aged to achieve better outcomes (green space, im-provements in air quality and sewage treatment) thanmost other developing-country cities.25

The growth of urban poverty in many countries,evidenced especially by the increase in populationsresiding in extremely poor environmental condi-tions, is partly a reflection of the pressures on lim-ited city resources.26 It also underscores the failuresof institutions and their unresponsiveness to certainconstituencies. As urbanization plays out, growthrates will slow, easing the pressure on cities. Butcatching up over several decades is not satisfactory.Measures are needed now to accelerate the rate ofimprovement in cities and to avoid making it morecostly to close the gaps later. There are now techno-logical and institutional options that need to be ex-plored more vigorously.

For cities to contribute to sustainable develop-ment, they need to maximize the positive while min-imizing the negative externalities. The favorable eco-nomic and social impact of cities can exceed their“ecological footprint.”27 Cities can get themselvesinto vicious or virtuous circles that become self- reinforcing, and triggers for change can often befound in institutional innovation. A key institutionalcatalyst—information—can increase urban benefitsand reduce diseconomies and risks.

Building informed constituencies to address

spillovers and anticipate risks

The spatial concentration of people and economicactivities in urban areas creates spillovers with signif-icant impacts on residents—and increasingly onwider regional and global populations. Mobilizingfor action to solve such problems (for example, pol-lution) requires that the parties affected gain accessto credible information on costs and benefits andthat they perceive a common interest in finding a solution. Building an effective constituency can bemore difficult where the impacts are uncertain andinfrequent, as in disaster mitigation. Advances intechnology and knowledge help, and local and na-

tional governments need to play important leader-ship roles in both cases.

Credible information and incentives—curbing air pollutionAir pollution generates large social and economiccosts. In many developing and transition countries,the damage reaches 4 to 6 percent of urban income,and has serious adverse affects on human health. Be-tween 500,000 and 1 million people die prematurelyevery year as a result of air pollution–induced respi-ratory problems.28 Vehicle emissions create the great-est damage to human health because they occur nearground level and in dense population centers, whilesmokestack sources disperse pollutants more widelyat higher elevations.29 Urban residents in low- andmiddle-income countries have greater exposure (wellabove WHO guidelines) to some localized air pollu-tants, such as suspended particulates, than their coun-terparts in high-income countries, even though thelatter consume more energy per capita.30

Countries do not have to suffer worsening air qual-ity as they industrialize, motorize, and become richer.Many technologies and behaviors for curbing urbanair pollution are cost-effective even at low levels ofeconomic development and limited institutional ca-pacity, as long as there is political commitment andpublic understanding.31 While action by industrialcountries to eliminate leaded gasoline, for example,took a decade to implement, sharing knowledge anddemonstrating workable solutions have permitted de-veloping countries to phase out this fuel much morerapidly (chapter 7).

Curbing stationary sources of urban air pollutants(concentrated interests) is institutionally easier thancurbing mobile sources (dispersed interests) becausethere are fewer polluters.32 That the fuel supply wasthe main source of airborne lead made it easier forcountries to implement the phaseout administra-tively. Pressure from an informed public has beenmore instrumental in getting governments to rein inother types of pollution and in motivating regula-tory or other action. In China the educated urbanpopulation has been an important force for such re-forms. But the national government’s willingness tomake information on health costs and risks publiclyavailable was an essential precondition.33

Curtailing mobile sources of pollution and largegas guzzling vehicles is most challenging because themiddle- and upper-income groups are the beneficia-

ries of increased motor vehicle travel, and the mainsource of growing emissions with global and regionalimpact. These stakeholders are a more influential in-terest group than the general public, and especiallymore so than poor people suffering from the result-ing pollution and accident risks. Collective action toreduce transport-based GHG, (especially CO2) isfurther complicated by the nonlocal and longer-termnature of the damages.

Growth in motor vehicle ownership in develop-ing countries could overwhelm improvements in fuelor vehicle efficiency. Effective approaches to reducetransport-based pollution therefore involve a rangeof interventions at different scales (local, national,and global), forming part of the integrated transportstrategies discussed below.34 These measures includeimproved information on levels and sources of, anddamages from, pollution, educational campaigns, in-centive systems (including taxation of vehicles andfuels), and technological measures, such as replacinghigh-mileage, heavy-polluting vehicles and installingcomputerized inspection and maintenance regimes.

It is also necessary to manage supply and demandacross transport modes, through better public trans-port, improved conditions for nonmotorized trans-port, traffic-management, traffic-calming measures,35

and road and vehicle user fees. These measures re-quire coordination between jurisdictions within ur-ban areas and across levels of government.36

Creating constituencies—for clean water and wastewater managementEven though inadequate neighborhood disposal ofwastewater has unavoidable negative impacts, solu-tions are often limited by weak organizational capac-ity for collective action. Technological innovation,spurred by a professional association of progressive-minded civil engineers, was behind the introductionof a low-cost approach of shallow, small-bore sewer-age networks in Brazil a decade ago. Compared withconventional systems, this condominial sewerage re-duces investment costs by half but requires a strongcommitment of households to maintain it collec-tively. Condominial sewerage systems have workedwell where this cooperation was sustained and par-ticipating households understood their responsibili-ties. But since it can be difficult to get individualhouseholds to commit to take part, a more organizedinstitutional arrangement, through community as-sociations that would contract for the maintenance

work, is being considered for any future extensionsof the system.37

While communities and NGOs can work to-gether to provide household and neighborhood san-itation facilities, the social costs are lower and bene-fits greater if disposal and treatment of wastewaterare citywide. Few cities in developing countries treattheir wastewater, contributing to the pollution ofdownstream water bodies and corruption of fragilecoastal environments. Because investment costs arehigh and impact areas can extend across many juris-dictions, solutions require cooperation among localgovernments and across levels of government.

Water pollution charges have been much less ef-fective in making municipalities cut their emissionsthan industrial and agrobased polluters. The Wa-tershed Cap and Charge System in Colombia hasinduced industrialists to reduce effluents, yet two-thirds of local governments, which create 70 percentof the water pollution, have been unresponsive.38

Rising public demand for improved water qualitywill be needed to press for such changes, aided by in-come growth and good water-system management toreduce waste and sustain net revenues, since treat-ment is very costly.

For both air and water pollution, public awarenessand citizen pressure on governments and pollutersare possibly the most important factors motivatingenvironmental reform in the face of entrenched in-terests or official indifference. Political democratiza-tion and the freedom to associate can be catalytic inmaking dispersed stakeholder groups see their com-mon concerns and collective strength (box 6.3).

Mobilizing dispersed interests to anticipateproblems—preventing and managing disasters 39

Urban disasters (natural and industrial) can causelarge loss of life and have enormous economic andfinancial costs. They are especially devastating topoor people, who often live and work in precariousconditions. The drive for mitigation increases as theeffects of disasters, and the costs of failures to takeaction, become more immediate and widely per-ceived. Institutions are needed that can motivateaction in advance of crisis and share the costs andbenefits of preventive measures among citizens in afair manner. Hazard mitigation requires improvingknowledge, building constituencies for risk reduc-tion, and strengthening institutions and partnershipsacross levels of government and the private sector.

Climate change and natural disasters are closelyrelated. Global warming, projected to raise sea levelsas much as 0.8 meter this century,40 is particularlythreatening to coastal cities—where most of themegacities in developing countries will be located by202541—as well as to small island states.42 Climate-related events directly affecting urban areas includefloods, mudslides, heat inversions, wind storms, andstorm surges. Much of the economic and human tollwill strike at the advancing frontier between the builtenvironment and nature in the cities of developingcountries, which also serve as reception areas for en-vironmental refugees.

Although earthquakes claim fewer lives thanweather-related events on an annual basis, between40 and 50 of the fastest-growing cities in developingcountries are in earthquake zones.43 Exposure tosuch hazards and the reduced ability to provide basicservices after a crisis can jeopardize a city’s attractive-ness as a business location.

Building knowledge. Knowledge about the hazardsmay be scant or absent, even among residents mostat risk, yet community awareness of physical hazardsis fundamental for mitigation efforts. Comprehen-sive vulnerability assessments using remote sensing,satellite imagery, risk and loss estimation modelingcan help document and reduce physical, social, andeconomic vulnerability. Changing physical infra-structure and innovative techniques for retrofittingbuildings can improve disaster prevention. So can“soft,” nonstructural methods—those that increase

hazard information, create new knowledge, build the capacity of institutions, and train and raise theawareness of decisionmakers and the communitiesat risk.

Estimating losses can make the financial case forpreparedness. Memphis, Tennessee, calculated a $0.5million cost for retrofitting water pumping stationsto be disaster resistant—compared with $17 millionto replace each pump and $1.4 million for each daythe system is out of service.44 But developing coun-tries rarely have well-documented, location-specific,and hazard-specific costing of hard and soft mitiga-tion measures. Even more rare is costing used sys-tematically for public education. Without an edu-cated constituency collective decisions on disasterpolicies are usually dominated by better-off mem-bers of the community. Their priorities can differgreatly from those of poor people, who risk a largershare of their assets in a disaster.45

Creating incentives and constituencies for risk reduc-tion. Indispensable for mitigation strategies arestrong disaster prevention proponents and the polit-ical will to lead regulatory changes and financial ap-propriations. With limited resources developingcountries must rely on partnerships of all actors. Inthe United States disaster prevention started withcoalitions of scientists, emergency relief organiza-tions, professional associations, and other civicgroups who lobbied governments to fund researchand hazard mitigation strategies. This movement re-ceived impetus when the Federal Emergency Man-

The city of Cubatão, in São Paulo State, Brazil, was castigatedin the country’s press in the late 1970s as the “valley of death”because of the extreme industrial pollution of its water, air, andsoils that had occurred under decades of military dictatorship.Poor people lived in the midst of toxic waste dumps in an areaalso prone to natural disasters. In 1983 the state environmen-tal protection agency (CETESB) initiated a pollution control pro-gram that significantly reduced pollution levels in less than 10years. The agency managed to challenge the privileged posi-tion of entrenched industrial interests and make industrialistsbear most of the costs of pollution control.

How was this achieved? A citizens’ association of “victimsof pollution and bad living conditions” (APVM) focused thepublic debate on the human toll of pollution and attracted wide-spread support for reform. While the program was advancedby high-level support for environmental improvement in thestate government and aided by international opinion, three

changes at the national level in the early 1980s were key: thetransition to democracy that allowed the emergence of inde-pendent social activism, free elections at the state level, andthe elimination of media censorship.

In executing the program, CETESB relied on a suitable leg-islative framework and its status as the only agency with boththe mandate and expertise for pollution control in the state.Still, the environmental clean-up of Cubatão was possible onlyafter changes in the political rules of the game meant that pro-gressive bureaucrats could ally with informed citizens to chal-lenge the powerful economic elite that had stymied previousreform efforts. The experience of APVM forged a collectiveidentity among victimized residents that moved them to actand set new terms for a collaborative relationship between thecitizens and economic powers of the city.

Source: De Mello Lemos (1998).

Box 6.3

Political reform and stakeholder alliances overturning pollution

agement Agency (FEMA), armed with a federalmandate and incentives, took the lead and promotedlocal and state initiatives (such as the regional Earth-quake Preparedness Projects in California), but stillworked through civic and professional partners.

The public needs to decide on acceptable levels ofrisk, comparing the immediate benefits of expendi-tures on other social priorities with the delayed ben-efits of reduced loss of life and asset replacement costfollowing a potential disaster. These tradeoffs can beeased when well-designed incentives change privatebehavior to help prevent hazards. Examples includereducing insurance premiums on residential prop-erty when basic hazard-resistant steps are taken, of-fering disaster insurance with strict enforcement of building code provisions, or providing tax holi-days or grants for mitigation.46 Poor residents, forwhom insurance or fiscally based incentives may notbe practical, would benefit from urban planning forslum prevention, enforceable environmental zon-ing in cities, and resettlement combined with com-munity-based upgrading and tenure regularizationschemes (discussed below).

Recent disasters can motivate countries to under-take some of these measures and instill longer-termthinking. Gujarat State in India is trying to establisheffective disaster management institutions followingthe January 2001 earthquake that killed 15,000 peo-ple. The state has a new disaster management author-ity to coordinate all aspects of the response, workingwith NGOs, the private sector, universities, local com-munities, and external donors. The program includespredisaster preparedness and postdisaster response, re-construction, and disaster prevention. Incentives arebeing introduced to build constituencies for disasterprevention by capitalizing on the population’s height-ened awareness and willingness to change.

Adapting to climate change. Adapting to climatechange may be more difficult since the risks mountgradually and less visibly—but no less urgently.47

Coastal cities and other population centers (espe-cially small island states) will need to invest in pro-tective barriers and possibly to relocate residencesand essential public facilities through managed re-treat. Priorities for such adaptation should be givento built areas and infrastructure that require urgentattention in any case, such as vulnerable informalsettlements and outgrown sanitation and drainagesystems. Adaptive expenditures will place a signif-icant burden on the public sector, private utility

companies, and, indirectly, on the urban economy. Low-income residents living in harm’s way will needparticular assistance.

Balancing interests to provide

urban public goods

Urban areas can enhance and enrich social integra-tion through the provision of public goods and cul-tural and environmental amenities. Achieving thesebenefits requires institutions to channel dispersed in-terests of a pluralistic public—to give expression tothe social value of equitable access to publicly pro-vided assets, and to identify future needs in land de-velopment and redevelopment—and sometimes toovercome powerful vested interests. To provide otherurban public goods, similar mechanisms are neededto balance competing interests (for a well-integratedtransport system, and sanitary solid waste disposal),48

and to express dispersed interests (for drainage).Foresight, political will, and a governance system ac-countable to a wide array of stakeholders are key in-gredients for achieving credible commitments.

Balancing private and public interests in land useand committing to priorities in the public interestThe challenge in anticipating urban populationgrowth is to focus on the most socially, environmen-tally, and economically important aspects of futureland uses, and commit to credibly executing thesepublic choices. New settlements in or near existingurban areas require the following actions:

� Setting aside rights of way for primary transportarteries

� Proscribing settlement or other development ofareas that are unsuitable because of environmentalfragility or vulnerability to disasters (steep hill-sides, flood zones)—and protecting fragile envi-ronmental resources (urban watersheds, wetlands)

� Reserving areas for amenities, especially parks, anddeveloping other public spaces with social and cul-tural value.

This effort has to take into account emerging sup-ply and demand and avoid overdetermining the city’sfuture. A frequent problem is that city master plansmay exclude large high-value sites, especially at theperiphery, from urban development, while failing tofence off environmentally vulnerable or risky sites inan enforceable way. Although urban expansion into

agricultural areas may pose real social welfare trade-offs, much so-called agricultural zoning around citiesis outdated. It neglects the greater economic and fis-cal benefits of urban land uses and fosters opportu-nities for corruption and speculation. Urban munici-palities often have much less say over land conversionat the periphery than do national governments orpowerful elites.

To identify minimal, high-priority, and enforce-able limits on land use, local institutions must firstidentify the socially desired outcomes for urban de-velopment (such as which environmentally sensitiveareas to protect—recall the Catskill example in chap-ter 2, and how much green space to set aside) andthen commit credibly to achieving them. In Conakry,Guinea, the municipality developed a basic structureplan for the city in the early 1980s. Thirteen yearsafter the plan was adopted to create the primary roadsand infrastructure networks essential for urban mo-bility and productivity most of the networks are inplace. Now the city’s focus is on upgrading denselypopulated neighborhoods by providing internal andsecondary roads linking to this network, and basicmunicipal services (drainage and solid waste collec-tion), as part of an integrated citywide program.

Urban redevelopment for public goods. The eco-nomic and social vitality of cities is enhanced bypromoting and protecting their cultural characters,developing public spaces and other amenities as eq-uitable social assets, and converting abandoned anddegraded land and waterfronts to new uses. Civicgroups and private entrepreneurs can motivate suchcollective action, but often formal partnership andpolitical leadership from local and national govern-ment are needed.

A city’s historical heritage and social culture em-bedded in its neighborhoods and structures are valu-able assets.49 Vision and voice for dispersed and fu-ture interests are necessary to give appropriate weightto such intangible values, to counteract pressures to rebuild and modernize for commercial or high-income uses, and to prevent the gradual degradationof the built environment because of poor house-holds’ need for affordable places to live. In the early1980s the city government of a historic Chinese city,Ping-yao, almost demolished the old city wall tobuild a wide road. Protests by scholars and residents,with help from officials and news media elsewherein China and abroad, persuaded the government tolocate new development outside the old city. The

city’s economic decline has been reversed and it nowenjoys a new distinction for tourism as the only cityin China preserved within old city walls.50

The use and reuse of public spaces provides anopportunity to cultivate a city’s natural resources—by preserving or creating parks, architecturally dis-tinctive streetscapes and squares, and waterfronts.Such urban amenities are part of a city’s portfolio ofassets and broad access to them enhances well-beingand strengthens social capital, since poor and wealthyalike can enjoy them. In the fast-growing cities of thedeveloping world reserving open spaces requiresstrong commitment (championship) and forwardthinking to speak for these dispersed interests. A for-mer mayor of Bogotá saw in the hundreds of kilo-meters of drainage canals crisscrossing the city a wayof connecting all parts of the city by converting themto walkways and bikeways. Despite resistance fromsome quarters, the city realized this vision, creatingone of the world’s largest pedestrian paths lined withtrees, lighting, open sculptures, and benches—andlinking some of the poorest neighborhoods of thecity with golf courses and parks.51

Remediation of despoiled sites and abandonedstructures (brownfields) can also present many ben-efits to a city. Overcoming the institutional andfinancial barriers to brownfield redevelopment isimportant to correct the environmental and socialblight in surrounding areas and to prevent the flightof new (greenfield) investments to the urban periph-ery. Extensive experience in industrial countries un-derscores the social and economic benefits whenbrownfields are redeveloped as an integral part ofneighborhood renewal processes.52 Public–privateinvestment partnerships require clear legal frame-works for property rights, risk sharing, and assign-ing liability for pollution clean-up. Efforts in Bu-dapest to identify a redevelopment program for theCsepel steel works on an island in the Danube Riverin the middle of the city have been delayed by diffi-culties in negotiating with more than 200 owners ofthe site, which was hastily privatized during the post-socialist transition.

Leadership and foresight are also key to urban re-newal. In the mid-1980s the mayor of Istanbul ledan initiative to clean up an inlet of the Bosphorus,known as the Golden Horn, which had become pu-trid from the dumping of sewage and solid waste.The program—part of a large metropolitan sewerageinvestment to extend sanitation to some 300,000

low-income residents and treat municipal water—also relocated the polluting industries and ware-houses. The water quality of the Golden Horn hasbeen restored. Recreational and tourist activity hasreemerged. And public and private investment has re-juvenated the area’s historic and cultural assets.53

Balancing competing interests for accessibility54

Urban transportation is closely tied to urban landdevelopment and can create both positive and nega-tive externalities as cities grow. Urban transport isbest addressed as part of integrated urban strategiesthat can attend to the interests of all user groups (including poor people, women, and the mobility-impaired) and anticipate long-term needs that haveno vocal constituency. Most cities in developing andtransition economies are sufficiently densely popu-lated to support extensive public transport,55 andoften include (depending on physical and climaticconditions) walking and cycling as major modes oftravel. It is important that institutions are developedthat balance and give weight to these interests. Andsince motorization is still at an early stage, it is likelythat an urban transport strategy will emerge in suchan institutional setting that focuses on balancingroads and private cars with other alternatives withina broader urban perspective.

Traffic congestion is one of the major negative ur-ban externalities. As cities grow and become richer,vehicle ownership and use increase more rapidly thanavailable road space. Expanding road space tends tostimulate more car use, so the imbalance continues.Experience shows that more road building is not theroute to a congestion-free future. Most important ishow the space devoted to roads is used and man-aged—for example, it should be organized hierarchi-cally to separate traffic flows for different purposes.Managing demand, through taxes and impact fees onroad construction and use, is politically much moredifficult once car dependency becomes entrenched.Cities need a minimal amount of space for circula-tion relative to their size to operate efficiently, andearly reservation of rights-of-way for major transportroutes is essential to good urban planning.56 Trans-port infrastructure costs rise sharply as cities becomemore densely developed, and the investments arelarge and “lumpy,” so requirements need to be con-sidered well in advance of actual demand.

Although the environmental and efficiency costs ofmotorization and related traffic congestion attract

vocal (and often competing) lobbies, the interests ofpoor people are less well expressed politically.57 Yetpoor people lose when the dominance of private vehi-cle traffic undermines support for public transportand space for nonmotorized options. Poor peoplebecome more restricted in their mobility—and aspedestrians, they suffer most from road accidents.High-speed roads often carve up low-income neigh-borhoods, increasing noise and ground-level pollu-tion. During the construction of the U.S. interstatehighway system in the 1950s, for example, the plan-ning criterion of selecting “least cost” sites caused in-trusion not only into environmentally sensitive areas(wetlands and so on) because their market land valuewas low, but also into the poorest urban neighbor-hoods, severing them from the rest of the city and has-tening their deterioration. The massive public spend-ing on the highway system, unbalanced by sustainedsupport for other transport modes, accelerated theflight to the suburbs of wealthier city residents and theeconomic decline of older inner-city areas, contribut-ing to persistent social problems in U.S. cities.

Urban transport strategies that focus on themobility of all residents, not just a few, to maketransportation more sustainable environmentally, so-cially, and economically contain a balanced array ofmeasures:

� Managing the existing road infrastructure to im-prove the traffic flow and calm speeds arounddensely populated areas

� Giving weight to the effects of induced traffic andimpacts on nonmotorized transport and the envi-ronment when evaluating new road projects

� Internalizing the social costs of road use by charg-ing land developers impact fees to finance newroads, introducing road congestion pricing (orsome proxy), and charging the full social costs ofparking

� Improving the viability of public transport by giv-ing priority to buses on restricted lanes, ensuringadequate financing, and improving operational ef-ficiency through regulated competition

� Protecting pedestrians and nonmotorized trans-port users by providing safe walkways and bicyclepaths

� Providing rail-based mass transit in very largecities with high transport demand, where it canalso serve low-income users, as in some LatinAmerican cities.

Planning and managing transport requires bal-ancing conflicting interests in an environment of un-certainty and risks. Many decisions have long-termimpacts and high costs, so coordination across thevarious modes of transport is needed. Cities thathave managed to execute a comprehensive transportstrategy, such as Curitiba, Bogotá, and Singapore,58

have combined political will and leadership withtechnical and professional competence (box 6.4).Similar integrated approaches are necessary to ensuretraffic safety.

Reaching a consensus and compensating losers—sanitary solid waste disposal The production of solid waste (including hazardouswaste) increases as economic activities shift fromagriculture to industry, incomes rise, and lifestyleschange. Its improper disposal can have environmen-tal consequences. Managing waste removal and dis-

posal is a coordination problem. Waste collection isusually the responsibility of municipal government,but in many cities the formal service covers at besthalf the waste generated. Informal private operatorsdo much of the pick-up, sorting, and recycling ofgarbage, and communities sometimes provide theservice for themselves.59 But these informal solutionsare rarely well integrated into a safe disposal system.Improper disposal of solid waste damages drainagesystems, contaminates groundwater, and releasesmethane, a potent greenhouse gas.

A key constraint to collective action and coordi-nation is the NIMBY syndrome (“not in my backyard”)—no community wants the waste disposalsite. Disposal is then neglected, or a facility is locatedwithout public discussion near the population withthe least political clout. Successes in public decision-making on solid waste disposal facilities (sanitarylandfills or incinerators) suggest several lessons:60

� First, public discussion should be early and open,with site selection based on transparent criteriaagreed on in advance by candidate communities.61

� Second, communities adversely affected should becompensated, through financial transfers or accessto other desired investments. In Canada and theUnited States the selected community typically re-ceives “host fees.”

� Third, the project sponsor needs to be credible inmeeting commitments to minimize environmen-tal impacts, through proper operation and man-agement. The facility should be monitored by thelocal community and local authority—and de-signed to retain functions for informal collectors,so that they can acquire less risky livelihoods.

The problem of safe disposal is particularly acutein large cities and metropolitan areas where severalmunicipal governments need to reach agreement on siting and cooperate to share costs and reacheconomies of scale. Many large cities have beenstymied by the absence of an appropriate governancearrangement. In Monterrey, Mexico, a special-pur-pose metropolitan authority runs a public companyto operate a sanitary landfill serving eight munici-palities. In Santiago, Chile, 14 communes created ajointly owned corporation to build and operate alandfill and gas recovery system through voluntarycooperation rather than a formal metropolitan au-

Since 1998 Bogotá has implemented a comprehensiveurban mobility strategy that includes promotion of nonmo-torized transport (bicycle paths), restriction of automobileuse during certain hours and days (approved by public ref-erendum), and a bus rapid transit system (Transmilenio).Using exclusive busways on central lanes of major roadsand a network of feeder buses and stations, the systemprovides express and local services and carries 45,000 pas-sengers an hour per direction. Vehicle operations, passen-ger access, and ticketing services are carried out by privatecompanies through competitively tendered concessions.The new bus system is attracting ridership by former carusers and restoring respect for public transport.* By mid-2001 the system had achieved high productivity (630,000trips per weekday) at a fare that fully covers operatingcosts, with no traffic fatalities. Some air pollutants havebeen reduced by 40 percent, and user travel time is downby 32 percent.

Bogotá’s transport strategy benefited from strongmayoral leadership in articulating a long-term vision andrepresenting the interests of noncar users despite the re-sistance of motor vehicle lobbies. The program also re-quired partnership between the private concessionairesand the municipal government, which financed and imple-mented the physical infrastructure and provided the dedi-cated road space.†

* Presentation by Peñalosa (2001), updated April 2002.† World Bank (2002a), box 8.2

Box 6.4

Meeting environmental, social, and economic

objectives through urban transport strategy

in Bogotá

thority. These examples show that collective actionby governments, as by other social entities, requiresgoodwill, trust, and a conviction that interests arefairly balanced (in this case, reflected in burden shar-ing of costs and environmental spillovers from thedisposal facility).62

Finding voice for dispersed interests—drainage Many cities also lack effective storm drainage sys-tems, and ill-planned construction closes off naturalwater courses. In Algiers, where massive flooding inNovember 2001 caused 800 deaths (700 in denselypopulated neighborhoods), a natural water runoffchannel in the city had been converted to a pavedroad. Overflowing of clogged storm drains and sew-ers during high rainfall is projected to become agreater source of disasters in major cities than riverflooding.63

The key institutional issue is that drainage has noclear constituency until major problems occur. Localgovernments may become motivated to act ondrainage issues when flooding affects the businessdistrict, as in Cabanatuan in the Philippines, wherethe local business community put pressure on themayor to invest in drainage infrastructure. In Kam-pala the local authorities had neglected for years to protect past investments in the Nakivubo chan-nel from settlement encroachment and obstructionwith solid waste. There, and in communities inEthiopia, recent reforms expanding local democracyhave raised the profile of drainage as a priority forpublic expenditure.

Inclusion and access to assets: Challenging the

institutional roots of urban slums

As noted in chapter 3, the evolution of good institu-tions to solve coordination problems is itself deter-mined by the extent of inclusion and access to assetsby a wider public. Access to urban land—the city’sscarce natural resource and most durable asset—iskey to a city’s economic, social, and environmentalsustainability. Institutions need to allow people tosettle securely, so that they can envision a future fortheir families and their city, while allowing for flexi-bility in land use. Informal, illegal, or quasilegalneighborhoods with seriously substandard livingconditions, often generically called slums, are an ob-vious manifestation of inequitable access to physicaland financial assets, to secure land tenure, and to po-

litical representation. They also reflect the failures ofgovernment to guide and facilitate the growth oflow-income housing and basic services for incomingmigrants through appropriate policy and planning.These communities grow through the enormous en-trepreneurial energy of residents who build the cityand provide its labor. With the right institutional en-vironment they can evolve more quickly into safe,healthful, and hospitable urban neighborhoods.

Geographical and environmental manifestations of exclusionPoverty and its many manifestations in cities can beappreciated only by looking at disaggregated (espe-cially spatially detailed) data. In Cali, for example,the incidence of income poverty is highest in periur-ban neighborhoods with precarious environmentaland infrastructural conditions. The eastern area (for-merly a lagoon) and the western zone (steeply sloped)are settled mainly by poor migrants and minoritiesliving in very crowded housing (figure 6.2).64

N

Low: < 20%

Middle: 21–40%

High: 41–50%

Very High: 51–70%

Ext. High: > 70%

6

7

9

8

2

13

45

18

2014

19

17

1110

13

16

15

12

Source: Departamento Administrativo de Planeación Municipal de Cali,Colombia, 1999.

Figure 6.2

Poverty in Cali, Colombia: 1999 headcount rates

Reducing the disparities in welfare among resi-dents within cities is one of the starkest challengesto a sustainable urban future.65 These disparities areoften masked by official data on access to water andsanitation, which do not accurately reflect problemswith the quantity, quality, and reliability of ser-vices—or the numbers of people sharing facilities indense settlements. In Accra, Ghana, for example,only 12 percent of the richest quintile of the popu-lation, but more than 66 percent of the poorestquintile, share one toilet or latrine with more than10 other households.66 Some 44 percent of house-holds in Mysore, India, have water connections, butonly 8 percent of those in informal settlements do.67

Such inequities help account for infant mortalityrates three or more times higher among the low-income households of many cities than among high-income households (figure 6.3).68 Poor people inAccra and São Paulo have higher death rates not onlyfrom communicable diseases of childhood but alsofrom respiratory and circulatory diseases and injuriesfrom traffic accidents and homicides. The threats af-fect all age groups, creating a web of insecurity.69

Local groups and agencies concerned with envi-ronmental health problems and epidemiology in

cities are combining census data and household sur-veys with a geographic-referenced information sys-tem to map service access and health outcomes byneighborhoods. In Porto Alegre, Brazil, for example,the local authorities use a detailed environmentalatlas for planning and management and for educa-tion in schools.70

The multiple environmental health and safetyrisks in urban areas are related largely to the condi-tions and location of settlement. Hundreds of mil-lions of urban dwellers have few affordable optionsother than to live on sites (usually public lands)where development has not been approved andwhere residents are therefore not officially entitled tourban services or protections. Such informal neigh-borhoods remain in squalid condition for decades.Since the home is also a major source of income(both from rental and home-based industry) and thehousehold’s main private asset, the social and eco-nomic burden of such physical conditions is pro-found. Many slums are also disaster-prone sites—onhillsides or floodplains, or near factories. Monsoonflooding in Mumbai claims hundreds of victimsamong the illegal occupants of hazardous areas—in-cluding the canals meant to drain the excess water.

The population of urban slums is estimated byU.N.-Habitat at 837 million in 2001. Based on1993 regional breakdowns, more than half are inAsia accounting for one-third of the region’s urbanpopulation. Slums house more than one-half of allurban residents in Africa and about one-quarter inLatin America and the Caribbean.71

Slum neighborhoods typically have dispropor-tionately high concentrations of low-income people(though not necessarily the extreme poor, such as thehomeless). They may also house middle-income res-idents in cities where formal provision of infrastruc-ture and housing markets are very weak. Residentsof inner-city slums, typically settled for many years,generally have better availability of infrastructure(though it is often poor in quality and unreliable).They also have more established communities andless physical isolation than residents of newer settle-ments, usually on the outskirts. Both groups sufferfrom the stigma of their neighborhood that impedestheir access to employment and to wider networksof social capital.72

Factors associated with crime and violence are alsocommon in zones of deprivation within cities. The

0

20

40

60

80

100

120

140

160

0 20

Urban infant mortality rate, second poorest income quintile

Urban infant mortality rate, richest income quintile

40 60 80 100 120 140 160

MLI

INDEGY

TURIDN

DOM BOL

NICBRA

PERPHL

COL

Figure 6.3

High inequality in health outcomes

in urban areas

Note: Data shown for all countries with data base for which adequatesize sample is available (lowest quintile inadequately represented to beshown). Infant mortality rates measured as deaths before one year ofage per 1,000 live births.Source: Demographic and Health Surveys (see www.measuredhs.com).

highest homicide rates in Cali are found in its poor-est neighborhoods. Surveys of urban residents inGuatemala and Colombia identified tensions overaccess to water as a cause of violence.73 Analysis ofadministrative regions or municipalities in São PauloState categorized their territorial exclusion based onphysical hazard, provision of urban services, and se-curity of tenure. Municipalities with the most pre-carious living conditions had the highest homiciderates, and those with the least territorial exclusionwere the least violent. The regions with the worstoutcomes also had very high income inequality.74

Such exclusion contributes to frustrated expecta-tions, defeated hopes, and mistrust in society’s future.

Even in some transition economies many of theurban poor live in severely substandard, quasilegalsettlements (for example, 15–25 percent of urban res-idents in the Former Yugoslav Republic of Macedo-nia, many of them ethnic minorities).75 But not all in-formal settlements feature low-quality housing. Someillegal or irregular housing is produced by commer-cial developers or politically influential parties whospeculate that property investments will be regular-ized later (akin to the race for land rights at the agri-cultural frontier described in chapter 5). Such landspeculation is encouraged when countries lack clearpolicies on tenure security, and authorities are unableto balance interests and articulate public choices re-garding land use, or commit to enforcing them.

Empowerment through access to assets: security of tenureAlthough slums reflect institutional failures in hous-ing policy, housing finance, urban planning, publicutilities, and local governance, one of the most fun-damental failures is the absence of tenure security.Security of tenure means “protection from involun-tary removal from land or residence except throughdue legal process.”76 The emphasis is thus on pre-venting forcible and arbitrary eviction, whether ofindividual households or entire settlements. The sig-nificance for urban poverty underlies its inclusionunder the Millennium Development Goals (MDGs)as indicator 31 (“Proportion of people with access tosecure tenure”), related to Target 11 (“By 2020, tohave achieved a significant improvement in the livesof at least 100 million slum dwellers”) and Goal 7(“Ensure Environmental Sustainability”). The issueis also a focus of the U.N.–Habitat Global Cam-

paign for Secure Tenure and the Cities withoutSlums program of the multidonor Cities Alliance.

Lack of secure tenure in urban areas has not beensystematically measured. Even the designation of res-idents as homeowners or tenants does not conveyprotection from summary eviction when land regis-tration and the rule of law are poorly enforced.77 Se-cure tenure is part of a country’s hierarchy of rights,ranging from legal titles and contracts to customaryrecognition of use rights.78 Providing secure tenuretherefore does not pit the rights of squatters or ten-ants against those of private property owners andlandlords, who should be protected under contractlaw. But countries permitting arbitrary eviction oftenalso fail to enforce private real estate contracts andotherwise obstruct the private rental market, furtherdisadvantaging low-income citizens.

By confirming the rights and responsibilities asso-ciated with the occupation and use of land, regulariz-ing tenure status removes a major source of economicand political insecurity for households and for com-munities. It reduces some of the risks that discourageresidents from investing in their houses and shops—and gives them a stronger stake in urban society andan incentive to work with local officials to obtain ser-vices. A study in Indonesia found that stronger tenuresecurity increased the probability of demandinggarbage collection.79 And surveys of slum dwellers inBangalore reveal that better tenure status has a signif-icant and positive impact on willingness to engage incollective action to obtain urban services, even in cul-turally heterogeneous communities.80

A growing commitment by city, state, and na-tional governments of Brazil to regularize slums orfavelas has put in train a process of transformation(box 6.5). A key turning point was the 1988 federalconstitution, which strengthened the role of localgovernment and encouraged municipal policies to le-galize and improve tenure conditions in these infor-mal settlements. A groundbreaking new city statuteenacted at the federal level in July 2001 provides alegal underpinning for municipalities to regularizefavelas as part of concerted plans to combat spatialsegregation and social inequity—and to create moreinclusive and democratic urban governance.81

Security of tenure is both a collective good and aprivate good in the urban context.82 Whole commu-nities are threatened when shanty towns are bull-dozed, while residents individually gain security

when their settlement is accepted as an integral partof the urban fabric. Often there is enough stabilityof inhabitants that they confirm each others’ rightsto residence. Community organizations in informalsettlements use their strongest asset (social networks)to protect their area from encroachment by new en-trants, to resist involuntary resettlement, and to push

for associated rights as urban citizens.83 The RailwaySlum Dwellers Federation in Mumbai (RSDF) suc-cessfully managed the resettlement of member house-holds to permit a transport project with citywidebenefits (box 6.6).

How important is a legal title? Experience in manydeveloping countries confirms that households and

In many cities in Brazil, large shares of the population—25 per-cent of the residents of Rio de Janiero and 40 percent in met-ropolitan Recife—live in informal or illegal settlements, oftenon public lands. These favelas are home to an essential work-force—a workforce subject to terrible health conditions, fre-quent natural disasters such as mudslides and floods, andcrime. Official policy toward favelas in the past was that of ne-glect (with occasional introduction of services when politicallyexpedient or necessitated by emergency) and threats of evic-tion. Not until the 1970s did most municipalities begin to eveninclude such settlements on planning maps as provisional, de-spite their existence for decades in many cases. Transforma-tion in these settlements has started to occur in recent yearswhere local governments, supported by their state and the na-tional government, have made commitments to sociopoliticalas well as physical inclusion of the favelas into the city.

Beginning in the early 1980s a number of cities, most no-tably Belo Horizonte and Recife, initiated efforts to regularizeor integrate the favelas into the urban fabric and give themlegal recognition. New planning instruments were introducedat the national level to permit designating certain settlementsas “special residential zones of social interest” (ZEIS), whichpermitted planning and zoning regulations to be adapted to theland use requirements of these communities. In Recife, a fur-ther mechanism (PREZEIS) was established in 1987 to institu-tionalize, for the first time, the process of integrating irregularsettlements into the formal planning apparatus, with commu-nity participation, and allow for the provision of services andinfrastructure to reduce disparities. Under this law Recife cre-ated a land tenure legalization commission charged to identifyand address specific problems in each ZEIS through participa-tion of multiple stakeholders—a device credited with enforc-ing the government’s commitment to follow through with itsregularization program despite resistance from conservativesectors. The state of Pernambuco has joined Recife’s effortsby bringing investment resources to help cover the settle-ments designated for regularization across the metropolitanarea.

Programs with similar objectives have been adopted inother Brazilian cities, including Porto Alegre, Rio de Janiero,and São Paulo. The Rio program (where the state governmenthas also reversed its past resistance to favelas, by providingfinance for building materials for residents without requiringcollateral) is notable for its scale.* In Belo Horizonte and PortoAlegre the programs entail a strong emphasis on participatorybudgeting and planning for investments in the settlements. A1998 study by the Brazilian Institute for Applied Economics in-dicates that at least 794 municipalities have some kind of

favela or informal settlement upgrading program and about506 of these include some form of land tenure regularization.

Where tenure regularization policies have intended totransfer full individual freehold titles to the occupiers of publicor private land, as in Belo Horizonte and Rio, this aspect of theprogram has been problematic to implement and less success-ful than the physical upgrading and service provision. Othermunicipalities, such as Porto Alegre and Recife, have used aninnovative alternative legal instrument to promote individualand community security of tenure. This formulation, the “con-cession of the real right to use” (CRRU), is a leasehold thatconfers private property rights to publicly owned land for a pe-riod of up to 50 years, either for an individual or a community.†

Combined with the designation of settlements as “zones ofsocial interest,” the CRRU protects residents from evictionand gives them broad property rights. This instrument permitsthe state to protect access for the low-income communitiesto land they occupy in order to promote socioeconomic inte-gration of the city; it also serves to preserve scarce public landfor current and future social uses. Settlements granted suchuse rights have gained physical improvements from privateand public investment in housing and infrastructure, and in-creasingly take on the appearance of working class neighbor-hoods physically integrated with the adjacent areas.

Though still largely untested, the new national City Statutegives municipalities the tools to go even further in regularizinginformal settlements. The Statute includes, for example, provi-sions for facilitating the transfer of privately owned land to theexisting occupiers in cases where occupation has gone uncon-tested for at least five years. To complete the transformation inthe quality of life and social inclusion of the urban poor, thesepolitical and legal commitments and investments need to besupplemented by a broader set of policies that also promoteeconomic opportunity and counter other dimensions of poverty.

* The first phase upgraded about 60 favelas and 20 irregular subdi-visions, affecting some 250,000 people, supported by an Inter-American Development Bank loan in 1995. A second phase and loanin 2000 applies to 56 favelas and 8 irregular subdivisions, for aboutthe same number of residents. The total costs of upgrading underthese projects average $4,000 per household, which comparesfavorably to most housing programs and to many sanitation andsocial services projects. Brakarz (2002).† In Brazil, the CRRU can be registered, allows transfer of right tolegal heirs, selling, renting and use of land as collateral. It can be usedas both individual and collective right (a form of condominium), andhas a gender dimension, as women are given priority treatment ingranting of use rights (Payne and Fernandes 2001; Fernandes 2002).Source: Cira, background note for the WDR 2003.

Box 6.5

Regularizing favelas in Brazil

communities realize significant benefits in movingfrom highly insecure tenure to de facto tenure (moresecure though not fully legal).84 Granting individuallegal titles (freehold) is sometimes expected to con-fer additional advantages, such as greater access tohousing loans and a more active housing market.85

But this requires supportive banking and real estateinstitutions, which have often not materialized be-cause of other limitations, including low incomesand the reluctance of lenders to finance incrementalhome improvements, the main way poor people pro-

vide their own housing.86 In the transition econo-mies of Europe and the Former Soviet Union, someof the incentive effects of housing privatization havebeen limited by the informational and financial im-pediments in the real estate markets.87

Titling can even worsen poor people’s overall ac-cess to affordable land and housing. That can hap-pen when titles are extended only to certain settle-ments, when tenants are forced out by higher rentsafter titling, or when slumdwellers are resettled intonew neighborhoods with deeds of ownership but noassistance in acquiring infrastructure services.88 Ef-fective demand for legal titles is therefore often lessthan would be expected by observed increases inland values.89

Legal titling can also have significant administra-tive costs, especially since many cities lack goodcadastres, leading to protracted legal battles overownership status. The COFOPRI program in Peruachieved a much higher volume of legal titling (onemillion deeds issued in four years) than did similarefforts in other countries. But this has been possiblebecause of the large tracts of government-ownedland on the urban periphery. Titling in older urbanareas of the country has progressed more slowly be-cause of ownership disputes.90 An alternative to for-mal titling and cadastres that can be quite effectiveand easy to implement to aid service delivery andacknowledge occupancy is street addressing—themapping and naming or numbering of streets andhomes in informal settlements. This system, used in15 West African countries, aids utilities in billing forservices and permits the simple taxation of plots.91

When residents in informal neighborhoods do notfear arbitrary eviction, they can devote their social cap-ital to negotiate claims for services with local govern-ment or utility companies and take collective actionto improve their settlements.92 In Pune, India, a resi-dents’ organization used similar techniques of com-munity mobilization and self-assessment as in Mum-bai to respond to their own demands for sanitation,after these had been long neglected by the municipalauthorities. The Pune slumdweller alliance (mainlythe women) assessed the needs of residents, managedthe construction with innovative designs to meet theneeds of different user groups, and set up effectivepayment and maintenance arrangements. Their ef-forts resulted in a record outturn of latrines in just afew years, serving about half of the city’s 1 millionslumdwellers.93 In Santo Domingo, the Dominican

In Mumbai, the commercial capital of India and home to12 million people, some 24,000 families have lived for al-most two decades along heavily traveled suburban raillines, with some huts hardly a meter from the tracks. Be-sides risking death and injury, these residents suffer froma near-total absence of basic services.

A project to improve the city’s traffic and transportationsystem required the resettlement of these slumdwellers.To represent civil society in the resettlement plan, the Ma-harashtra Government task force sought the participationof an alliance of The Society for the Promotion of Area Re-source Centers (SPARC; a registered NGO), the NationalSlum Dwellers Federation (NSDF), and a savings coopera-tive of women slum and pavement dwellers. A constituentunit of the NSDF is the RSDF, made up of the Mumbaifamilies who would have to move for the railway project.

By June 2001 the alliance had resettled 10,000 families,in just over a year, without force, to accommodations withassurance of secure tenure and basic amenities of water,sanitation, and electricity. How was this done? The Mum-bai Metropolitan Regional Development Authority in chargeof the railway project was willing to give up some of thepowers normally held by government agencies in resettle-ment and rehabilitation—determining eligibility, obtainingbaseline information on the community, allocating housing.Such functions, which provide opportunities for rent-seek-ing and corruption, were ceded to the NGO alliance.

Long before the railway project was initiated, the RSDFhad collected information on the railway dwellers as ameans of community mobilization and had the trust of itsown members as a resource for the resettlement process.The households agreed on the criteria for allocating perma-nent and temporary accommodations. In the new settle-ments the families have formed lending cooperatives tocompensate for income forgone as a result of the move.

The experience shows that a mobilized and self-gov-erning community of poor people can act collectively forits own good and for the good of the larger urban societywhen there is mutual trust and flexibility on the part of thecommunity and government agencies.

Source: Burra (2001a).

Box 6.6

How railway dwellers in Mumbai managed their

own resettlement

Republic, residents’ associations of three low-incomesettlements took similar steps. These settlements, lo-cated on the sides of ravines are vulnerable to frequentlandsides. Association members, mainly the women,designed and managed their own disaster mitigationprogram by building retaining walls and making otherinfrastructure improvements.94

Institutions for sustainable urban development

Good urban governance requires institutions to re-veal and balance divergent interests and to committo solutions for collective welfare. Some institutionalarrangements are particularly important to ensureperformance of these functions across the range ofurban issues:

� A structure of responsibility sharing and coordi-nation that links the community, local govern-ment, and provincial and national levels of gov-ernment and empowers the appropriate actors toaddress problems at each level

� A forum for wide participation in strategic think-ing, to enable common understanding and con-sensus, motivate actions, and assess progress

� Networks for communication and capacity build-ing among practitioners and stakeholders.

As noted, some informal arrangements that workin rural areas can be paralleled in urban communi-ties, especially in neighborhoods. But such informalinstitutions are often asked to do too much. Theycan stimulate private enterprise but not enable firmsto grow. And they can provide support to house-holds but not confer all the services, economic secu-rity, or political legitimacy that the population de-serves at the scale the city requires. The greater scaleand complexity of urban life thus require effectiveformal institutions that operate with greater pre-dictability, transparency, and adaptability.

What prompts such institutions to emerge andgrow? Often major changes seem to come aboutthrough sudden crises—a disease outbreak or nat-ural disaster—or the rise of a charismatic leader. Inrecent years democratization and fiscal decentraliza-tion have given new legitimacy and authority to localgovernments (chapter 7).95 And globalization hasbeen creating new opportunities, bringing in newknowledge, and raising new expectations for address-ing urban challenges.

Empowering the appropriate level of actors and ensuring coordinationThe environmental and social assets needed for sus-tainable urban development get more complex withincreasing scale of settlement. Moving from neigh-borhood to city, to region, and to nation impliesmore extensive environmental and social linkagesand impacts, increasing the divergence of interestsand potential for conflict, and greater technical andinstitutional requirements for coordinating those di-vergent interests.

Taking responsibility for urban services and spill-overs at the lowest practical level—the principle of subsidiarity—is a basic condition for mobilizingcollective action. Subsidiarity empowers those withmost at stake and strengthens the legitimacy of highergovernment through power sharing. Decentralizingurban services to local government is desirable forenhancing the voice of urban citizens and the accessto credible information, but it requires the respec-tive authorities to be accountable and have themeans to address problems at their level—and thatdepends on the framework of intergovernmental fi-nancial relationships.96

Political tensions across levels of government arecommon. Central governments frequently imposeunfunded mandates on local governments, and localgovernments may innovate without getting adequatesupport from central governments. Local govern-ments can also be less progressive than central gov-ernments and obstruct needed reforms. In China,some local governments resisted disseminating in-formation on city environmental conditions wellafter the State Environmental Protection Agency hadauthorized it, while others (in Jiangsu province)moved on disclosure policies—experimentally—inadvance of the central government’s commitment.97

Collective action in neighborhoods. Local environ-mental problems, such as removing solid waste fromthe neighborhood, can often be addressed by coordi-nation at the community level. The commonality ofinterests makes collective action possible either tosolve problems internally or to obtain what is neededfrom other parties (government or utilities). Grass-roots and community organizations that move be-yond confrontation to engagement with city govern-ment have greater prospects for obtaining sustainablebenefits for the urban poor.98 Experience in Pakistan,the Philippines, and Thailand with programs for

basic infrastructure and housing improvements inlow-income neighborhoods demonstrates what com-munity associations can achieve as key players inpartnership with government and the private sector.Those partnerships, however, require a long-termcommitment to sustain them.99

The main weakness in many such programs hasbeen the lack of continuity in financial and politicalsupport from formal institutions. In particular, cityagencies and utilities need to assume responsibilityfor scaling up and maintaining infrastructure net-works and such services as drainage, lighting, andparks beyond the neighborhood. Links with off-siteinfrastructure (roads or solid waste disposal sites) areoften inadequate because of weak coordination withresponsible government agencies, reducing the envi-ronmental and other benefits.

Many urban neighborhoods have also used theirinternal social networks to create protective institu-tions to ensure local public safety, such as throughcommunity “crime watch” activities—but this alsorequires the close collaboration and support of for-mal institutions, such as the police. In the WarwickJunction district of Durban, South Africa, 50 tradersfrom the community voluntarily patrol around theclock. The group was trained by metropolitan policeon citizen’s arrests, constitutional rights of individu-als, and court procedures to ensure successful prose-cutions. Their efforts have contributed to a reduc-tion in crime, and new trust between the communityand police has improved the rate of successful policeinvestigations and prosecutions. But relations arestill strained by the citizens’ perceptions of inade-quate formal policing in the city.100

Strong local (municipal) government. Most issues of sustainable development in urban areas extendbeyond an individual neighborhood and require apermanent formal mechanism of collective action,through effective local government that works withcommunities but shoulders broader responsibilities.This is where clashes of interests among larger andmore powerful stakeholder groups—and tradeoffsabout priorities and who gets served—begin to occur,underscoring the need for representative governance.

Reforms to make municipal governments andtheir agencies more accountable and transparent toall constituents are thus at the core of improving sus-tainable development in cities. The reform agendacomprises:

� Increased democratization (electoral processes)� Good practices and incentives for sound financial

management� Public participation and access to information in

budgeting and investment planning� Upgrading of skills and professionalism of govern-

ment staff� Monitoring and evaluation based on benchmark-

ing and client feedback.

Many cities demonstrate innovative relationshipsbetween civil society and local government to in-crease pressures for performance in the execution ofbasic functions. Reform initiatives often gain groundthrough collaboration on concrete activities of thelocal government—such as the public review of mu-nicipal budgeting and procurement in Obninsk,Russia,101 and the citywide referendum to affirm res-idents’ willingness to pay for infrastructure improve-ments after flooding in Tijuana, Mexico.102 Such ini-tiatives can then launch wider, deeper processes ofreform in other areas. Obninsk is now influencingreforms at the provincial level, and Tijuana trans-formed its disaster recovery plan into a series of in-novations linking taxation and public work improve-ments. Participatory budgeting, which was initiatedin Porto Alegre and has now spread to over 80 citiesin Brazil, has dramatically transformed relationshipsbetween civil society and local government.103

As part of urban capacity building and decentral-ization in Senegal and Guinea some local govern-ments have institutionalized public consultation inthe production of their public investment and main-tenance programs and embedded these programs ina contractual agreement between the city and thecentral government. This municipal contract com-mits the authorities at both levels to enforcing theagreed financial implications and reform measurespromised to constituents.104

Collaboration across jurisdictions—metropolitanmanagement.105 The wide impact areas of many en-vironmental spillovers, and the interdependence ofmajor economic activities sharing spatially contigu-ous infrastructure networks and other services, re-quire collaboration among local governments, as wellas regional and national governments. This is espe-cially true for systems implying economies of scale—citywide transportation, water resources manage-ment, pollution control, landfills, and wastewater

treatment. But this next level in the hierarchy of col-lective action presents an even greater potential di-vergence of interests, especially over sharing costs andbenefits. And fragmentation and excessive competi-tion among municipalities are more the norm. In themore than 70 countries undergoing decentralization,with the notable exception of India, municipal lawsand constitutional reforms have had the largely un-intended effect of weakening the prospects for met-ropolitan solutions to large city problems.

A variety of organizational arrangements—formaland informal, in developing and industrial coun-tries—have evolved over the years to meet the chal-lenges of managing large cities. In one common pat-tern (as in Dhaka and São Paulo), jurisdiction forspecific functions is assigned by geographic area, cre-ating many general purpose local governments thatmay cooperate for specific purposes, such as gather-ing regional data or sharing the costs of expensiveequipment and facilities. Sometimes this collabora-tion results in a specialized metropolitan area or dis-trict service, in which limited powers—usually forplanning or preservation, as of watersheds or re-gional parks—are ceded to special authorities.

Formally constituted metropolitan authorities,created or authorized by national law, are less fre-quent. In a pattern called functional fragmentation,lower-tier governments are limited to basic functionssuch as water distribution and street lights, while asecond tier handles areawide functions, such as free-ways and water trunk supply lines, as in Mexico City(box 6.7). These second-tier governments can be au-tonomous bodies, sometimes with executive powersto carry out projects such as the development author-ity in Kolkata. In other cases, as was common inLatin America in the 1970s and 1980s, the metro-politan agencies have only a consultative role. Themost formal arrangement—that of centralized met-ropolitan organizations as in Bangkok, Kuala Lum-pur, and Seoul—is fairly rare and usually imposed bycentral government to manage a capital city.

Some cities, such as Johannesburg, have blendedthese models—and the original prototypes are alsochanging as demands shift. Globalization exposescities to forces requiring a wider basis for manage-ment and planning—and more strategic direction.Metropolitan arrangements also seem to be respond-ing to the need to connect voter-taxpayers with mech-anisms of public choice. In the past few decades, Lon-don, Montreal, New York, Ottawa, and Toronto have

all followed an iterative path, moving first (in the late1980s) from formal metropolitan bodies with execu-tive authority to a system of fragmented, independentmunicipalities. By the end of the 1990s these citieshad shifted back to various centralized systems, butwith more democratic input from elected or appointedcitizen groups.

As these and other arrangements evolve, interna-tional sharing of experience will be important. Lead-ership from national or provincial governments isoften necessary to provide functional assignmentsand funding for metropolitan arrangements, sincelocal governments do not cede powers easily, espe-cially when it involves redistributing tax revenuesacross jurisdictions.106

Building consensus and strategies for sustainableurban developmentAn essential condition for coordination is having ashared understanding of the problem, knowing thecosts and benefits of alternative solutions, and ac-

Mexico City has explored several organizational arrange-ments in recent decades. The urban agglomeration in theValley of Mexico starts with the Federal District, the seatof the central government, an area with 10 million peopledivided into 16 boroughs or delegations. Surrounding theFederal District, but seamlessly connected in functionalterms, are 12 contiguous municipalities (another 7 millionpeople) in the State of Mexico. An even more encompass-ing definition (the Mexico City Region) covers nearly 100municipalities in five states.

Various metropolitan commissions have been createdto cover key areas of need. A water commission wasformed more than 30 years ago to plan and implement amammoth system of interbasin transfers to supply waterto Mexico City. Similarly, an air quality commission hassteadily grown to manage mobile sources of pollution. Thecommission has successfully eliminated lead from fueland is working to reduce the volume of traffic and to im-prove vehicle operating efficiency.

A persistent problem, despite half a dozen planningand special purpose commissions, is coordinating thecity’s growth with water and transport infrastructure—forexample, having mass transit connect residential areas tokey concentrations of employment. Recent political re-forms making the office of the mayor an elected positionhave sparked an active political debate about creating anentirely new metropolitan authority for Mexico City.

Source: Campbell, background note for the WDR 2003.

Box 6.7

Mexico City’s search for metropolitan

management arrangements

commodating the concerns of different stakehold-ers. Two relevant sources of experience in revealinginterests and reaching a consensus on what to do arelocal environmental action plans and city develop-ment strategies.

Since the Earth Summit in 1992 some 6,400 localauthorities in 113 countries have either formallycommitted or actively undertaken to produce localenvironmental action plans (called “Local Agenda21s”).107 These initiatives integrate environmentalobjectives into development plans, emphasizing par-ticipation and accountability. They articulate localconcerns and motivate local stakeholders aroundshared priorities for the area’s future. And they pro-vide a basis for coordinating the work of differentlevels of government and sectoral agencies.

The city of Manizales, Colombia, with about360,000 inhabitants, has formulated a local environ-mental action plan with wide consultation and inte-grated it into the municipal development plan andbudget. The plan included measures to revitalize thecity’s architectural heritage, improve public transport,strengthen watershed management of the ChinchinaRiver which provides water to the city aqueduct, re-duce risks of landslides, create ecoparks, and definecommunity environmental action plans. The planalso devised an innovative indicators program of“environmental traffic lights” to signal progress.108

Not all Agenda 21 exercises have been as partici-patory as planned, and the momentum for longer-term implementation has often waned—the dif-ficulty of forging long-term commitments being oneof the barriers to successful coordination noted inchapter 3. Success has been greater with sustainedsupport by successive mayoral administrations, strongparticipation by a local NGO or university, and ef-forts to mobilize funding for local economic develop-ment. Most Local Agenda 21s have been undertakenin smaller cities, perhaps indicating the difficulty ofconsensus-building at a metropolitan scale. Nationalgovernment leadership has also been instrumental inreplicating Local Agenda 21s within countries.

City development strategies are similar efforts inparticipatory strategic planning, but with a poten-tially broader focus of integrating environmental andproductivity concerns within a propoor perspec-tive.109 They have been used to build consensus on a vision for the city and on the steps to achieve thevision. These strategies typically include a participa-tory assessment of the city’s economic, social, and en-

vironmental conditions and prospects, and spell outpriorities and action plans for both policy and invest-ment. Some examples:

� A city development strategy in Cali, Colombia,helped explore stakeholder views on major publicprojects and led to changes in the city’s investmentpriorities. The first phase, which also prompteddiscussion of violence as a key issue for the city,pointed out the need for better understanding ofthe local economy and employment constraints.

� In San Fernando, in the Philippines, a city devel-opment strategy helped reorder investment prior-ities for sanitation, among other outcomes.

� Santo Andre in São Paulo State focused its strate-gic planning on social inequities and exclusion—and undertook social mapping to target and mon-itor actions to reduce disparities.

A good practice for institutionalizing city strate-gies is to incorporate their key elements into the reg-ular systems of city planning. While sustaining andimplementing strategic planning remains a chal-lenge, such efforts can reveal the priorities of variousstakeholders and contribute to public pressure forchange.110 Two items that should be on the agendaof most city development strategies are: Getting aheadof the expanding urban frontier; and making urbandensity affordable and livable.

Getting ahead of the expanding urban frontier—guiding new settlements to prevent future slums. Citiesand towns in developing countries will need to ac-commodate the projected doubling of urban pop-ulation over the next generation. Even with insti-tutional reforms to upgrade and integrate existingslums, new ones may form. Local governments haveoften shied away from acknowledging the need toanticipate and facilitate the growth of low-incomesettlements, instead letting them fend for them-selves. Providing infrastructure networks after thefact is much more costly, however, especially for verydense settlements with irregular layouts or requiringresettlement. In Bogotá the urban developmentagency estimates that installing drainage networks isabout three times more expensive in informal settle-ments than in planned neighborhoods.111

Political will is essential to create an institutionalenvironment that senses and anticipates the de-mands from new entrants and permits forwardthinking and partnership among government, pri-

vate investors, and households. Valuable experiencein planning low-income settlements has come fromsites and services programs, usually initiated by localgovernment or its agents, to provide basic plot lay-out and minimal infrastructure (such as core sani-tary facilities) in advance of spontaneous devel-opment. Such a program in Lima aimed to preventthe growth of squatter areas by anticipating demand(box 6.8).

Forging long-term commitments is key to suc-cessfully getting ahead of the frontier. A strong sup-ply response is important in making such schemessustainable. Many well-meaning programs have beenstymied by the lack of affordable land and hous-ing—even for the middle-income groups. This hasreduced political support for minimal design/mini-mal cost approaches, and poor people have beenpushed to the end of the service queue. In Conakrythe municipality’s basic structure plan aimed in partto set aside periurban areas for new settlements andto test public-private partnerships for the productionof serviced plots. The government planned to facili-tate connection of the sites to infrastructure net-works while the private investors, through paymentof an “equipment contribution,” were to ensure

replication of the scheme. But the authorities couldnot commit to fully implementing this part of theplan when no private developers joined in, and re-sulting delays led to land disputes as squatters movedinto the area. In El Salvador, however, a private com-mercial company (ARGOZ) has carried out a prof-itable land development scheme for low-incomehouseholds for more than 25 years, with the help ofa conducive legal framework for urban land conver-sion and a determination to keep design standardsaffordable.112

Making urban density affordable and livable. Ac-commodating the growth of population in cities willinvolve both physical expansion at the peripheryand, in many cases, increased density of settlementin the city. Average urban densities are already vastlygreater in major cities in developing countries, espe-cially in Asia, than in North America and Europe—for example, Mumbai has almost 400 persons perhectare and Shanghai 500, compared with about170 for the Barcelona metropolitan area, and 40 forNew York.113 The key issue for mature cities of de-veloping countries is to provide the infrastructureand services (especially sanitation, public transport,and green spaces such as parks and playgrounds)

Lima, Peru, 1977—As we stood on a hill overlooking a new set-tlement on the edge of a vast desert plain about 10 kilometersnortheast of downtown Lima, the boy, about 7 or 8 years old,said he was an engineer. An engineer? “Well,” he said, “I amhelping the people draw the lines for the lots and build our bar-rio.” The settlement, which consisted of little more than chalklines and shanties of woven reeds and plastic sheets, wasgrowing day by day, as new arrivals were trucked in by the Na-tional System for Social Mobilization (SINAMOS). The settlershelped the surveyors lay out the plots and clear areas for play-grounds and community facilities.

The core group of settlers had initially organized itself tosquat on public land under high-tension power lines. They hadcome mostly from slums in the city center, where they rentedrooms or lived with family. On the agreed-on night, walking insmall groups and with plastic sheets wrapped around theirbodies, they converged on the selected site and built theirtents and shacks overnight. Having established themselves asa squatter settlement, they knew SINAMOS would removethem to a permanent, if unserviced, settlement, one of thenew pueblos jóvenes, or young towns.

SINAMOS, staffed mostly by young and deeply committedengineers, architects, and social workers, was established inthe mid-1970s. Armed with enthusiasm and an understanding

of the dynamics of settlements that John F.C. Turner had de-veloped in Arequipa and Lima, SINAMOS set out to meet thechallenge of rapidly growing squatter settlements on the out-skirts of Lima.* They called the low-income communities pue-blos jóvenes, which gave the squatter settlements a positiveimage.

SINAMOS developed a two-pronged approach. It upgradedexisting pueblos jóvenes, relying on community participationand relatively little public investment. Then, as squatter settle-ments continued to appear, it began a massive slum preven-tion program, providing surveyed plots to meet anticipated de-mand for low-income housing. The first minimal structure planwas gradually detailed down to neighborhood layouts and en-gineering studies for trunk infrastructure.

Twenty years later the pueblos jóvenes had become well-consolidated, low-income and low-middle-income neighbor-hoods with most urban services, schools, clinics, markets, andother amenities. The key to the program’s success was pick-ing up signals of demand, balancing interests, and committingto implementation—matching social coordination with the as-pirations of poor people.

* Turner and Fichter (1972).Source: Chavez, background note for the WDR 2003.

Box 6.8

Leading the advance on urban settlement growth in Lima

needed to make already high densities livable and ef-ficient, with ease of access to homes, work, and otherplaces. The challenge for cities that have not yetfilled in is to avoid making such spatial developmentunaffordable to their growing population.

Governments often try to control city size and in-fluence the spatial form of city growth by regulatingland use—through rules on minimum plot sizes androad widths, for example. It is common for cities topresent inconsistent signals to investors: an officialstance extolling compact urban form, countered byregulations and financial practices promoting lowdensity land uses that favor middle- and higher-income groups. At the start of the economic transi-tion in Cracow, Poland, municipal officials advo-cated higher density development of the inner city—yet rigid zoning persisted, inconsistent with both theplanners’ intentions and the market incentives.114

When land-use regulations aiming to limit densi-ties are strictly enforced, as in Brasilia, they drive upthe cost of inner-city housing and force poor peopleto the periphery, where infrastructure services andtransport are unavailable or expensive.115 Even citiesthat try to encourage development around publictransport zones (such as Curitiba, Brazil) or to curbperiurban expansion by imposing a green belt (suchas Portland, Oregon) have seen denser land uses shiftto the outskirts. In Mexico, the lack of financing forrehabilitation of existing housing or for most multi-family units, as well as vestiges of rent control, deterimprovement of inner-city neighborhoods and pro-mote development on the urban periphery.116

Urban growth controls aimed at tightly regulatingdensities and building codes can make access to urbanassets of land and housing more inequitable.117 Zon-ing that permits mixed land uses—consistent withhow low-income neighborhoods develop naturally—is more advisable to keep jobs, services, and housingaccessible. Fiscal and other policies that charge devel-opers the full costs of providing the incremental in-frastructure required for new settlements are also nec-essary to internalize the social cost of expanding urbandevelopment. These charges, which should be intro-duced before spatial expansion patterns are locked in,can be a combination of development impact fees andgeneral taxation linked to property values.

But such charges will not reduce the demand forlow-income settlements at the periphery in develop-ing-country cities. Residents in these areas are notsubject to formal taxation as long as they lack tenure

security, and they already pay dearly for whatever in-frastructure they are able to acquire from informalmarkets. So, regularizing informal settlements and fa-cilitating low-cost land and housing developmentshould be the highest priorities to ensure more equi-table access to urban assets and more healthful, attrac-tive living conditions in developing-country cities.These measures, coupled with appropriate allocationof urban land for public purposes—right of way, en-vironmental easements, and so on—can transformthe institutional basis for the evolution of urban form.

Promoting institutional learning and leadershipthrough networksInstitutions for sustainable urban development needto embody incentives and processes for learning tobetter solve existing problems, and anticipate andprepare to deal with new problems. Increasingly thisstimulus occurs through networking by local govern-ment and nongovernmental groups, through bothassociations and Web-based communications. Net-works foster communication among peers, dissemi-nate innovations, and enhance reputational pressuresfor change. They can also instill professionalism andhigh standards of performance, cultivating leader-ship. Increasing forums for public feedback also helpsidentify mistakes and make mid-course corrections.

A prime illustration of networking is the asso-ciation of local, national, and international NGOspromoting the empowerment of communities andwomen in India and other countries. Through thealliance of SPARC, Mahila Milan (a women’s savingscooperation), NSDF of India, and Slum DwellersInternational (see box 6.6), practitioners and theurban poor share experiences on housing, urban ser-vices, and security of tenure. And by distributing in-formation gathering methods (such as a self-censusof slumdwellers), negotiating skills, and encourage-ment across the city and country and communicat-ing with similar groups elsewhere, the alliance is in-creasing the scale and sustainability of its efforts.

Local governments are also networking interna-tionally to learn from each other. National, regional,and international associations of local governmentsdiffuse technical assistance, training, and ideas tomember cities on a wide range of planning, opera-tional, and fiscal issues.118 The Union of CapitalCities of Ibero-America (UCCI) runs workshops and a Web site to help member cities learn aboutmunicipal modernization, solid waste management,

urban transport, cultural heritage protection, andother issues.119 The first cohort of Philippine citiesthat carried out city development strategies is help-ing others do the same, as part of a growing urbanknowledge network involving the Philippine Leagueof Cities, the China Association of Mayors, andother national groups in East Asia.120 The Clean AirInitiative, a consortium of donor and private fund-ing, is helping build capacity among cities in severalregions—for example, to extend to African cities theAsian and Latin American experiences with remov-ing lead-based fuel (chapters 3 and 7).

Networks also help to create incentives for sus-tained collective action by building reputational pres-sures within peer groups. Professional associations oflocal governments provide advice and standards onperformance indicators that can be compared orbenchmarked among member cities. Numerous ex-ternal rankings of cities on quality of life or attrac-tiveness to investors have been widely publicized—and have sometimes provoked corrective action.121

Conclusion

Achieving sustainable urban development requiresforward-looking institutions that sense emergingproblems, balance interests (especially by heedingthe disadvantaged in society), commit to effectiveexecution of agreed solutions, learn, and adapt. Tomake such institutions emerge and function well, itis necessary to confront basic inequities in access toassets, to empower dispersed interests and balancethem against vested interests, and to build consti-tuencies that can represent and commit to longer-term concerns.

Priority actions to reveal problems and divergentinterests include developing disaggregated datasets,such as mapping environmental hazards within acity. Wide dissemination of such information, alongwith the costs and benefits of alternative solutions,is essential to building constituencies for action. Bal-ancing interests and forging consensus can be facili-tated through participatory strategic planning, aidedby networking among practitioners and local gov-ernments for sharing local and global knowledge, in-novations, and reputational pressures that stimulateleadership. These measures to strengthen the work-ings of institutions are fairly low in cost and can beimplemented in the short term with existing capaci-ties and resources. Support from the central govern-

ment can help, but city stakeholders should takemuch of the initiative.

More fundamental changes would have wider andmore lasting impact—and would be more instru-mental in building new institutions. These deeperreforms include granting secure tenure, which cantransform the balance of power between urban poorpeople and the rest of the urban society. Increasingthe openness and accountability of local govern-ment, through democratic processes and partici-patory procedures, would also increase the respon-siveness to the interests and problems of poorerconstituencies and the legitimacy of governmentactions. Although these reforms are long in impact,experience shows that they can be initiated fairlyquickly—provided there is political will. More com-plex measures—such as devising metropolitan man-agement arrangements, and helping cities mitigatedisaster risks and adapt to threats from climatechange—may require more creativity, leadership,and resources. They also require greater and sus-tained support from national institutions.

Much of the future physical development thatcities require can come at a lower cost when problemsare recognized sooner rather than later; for example,setting aside rights-of-way for primary transport routesand parks and green spaces, facilitating new low-costsettlement, and guiding land development away fromprecarious or environmentally fragile areas. Invest-ments to protect environmental health locally can alsobe made effectively and cheaply with the participationof residents to identify and carry out appropriate so-lutions. Significant advances in the quality of life ofthe less advantaged urban residents are possible whenthere is a shared commitment to integrating themfully into the life of the city; when there is flexibilityon the part of government and private service pro-viders; and when there is an openness by formal insti-tutions to creative solutions developed by a diversearray of residents, actors, and networks in a city. Manyactivities in one location or community have conse-quences that affect other locations or communities.The principle of subsidiarity requires that these spill-overs be addressed at higher levels—a principle basedon matching the span of the spillover with the spanof the jurisdiction best able to internalize the prob-lem. The principle of inclusion ensures that people’swell-being is a priority to be addressed at national andglobal levels, as discussed in the next chapters.