Governance Failure: Rethinking the Institutional …...governance reform, and by raising the question of whether household provision of networked water supply by monopolistic providers

World Development Vol. 36, No. 10, pp. 1891–1915, 2008� 2008 Elsevier Ltd. All rights reserved

0305-750X/$ - see front matter

doi:10.1016/j.worlddev.2007.09.015www.elsevier.com/locate/worlddev

Governance Failure:

Rethinking the Institutional Dimensions of

Urban Water Supply to Poor Households

KAREN BAKKER, MICHELLE KOOY *

University of British Columbia, Canada

NUR ENDAH SHOFIANI and ERNST-JAN MARTIJN *

ESP-USAID, Jakarta, Indonesia

Summary. — This paper applies a conceptual framework of ‘‘governance failure’’ to an analysis ofthe institutional dimensions of urban water supply provision to poor households, focusing on thecase of Jakarta. Data from a household survey, archives, GIS-based mapping, and interviews areused to document governance failures that create disincentives for utilities to connect poor house-holds and for poor households to connect. The paper concludes by suggesting that the debate overthe relative merits of public and private provision has diverted attention from the pressing issue ofgovernance reform, and by raising the question of whether household provision of networked watersupply by monopolistic providers (whether public or private) is universally feasible given the cur-rent water supply policy norms.� 2008 Elsevier Ltd. All rights reserved.

Key words — water, infrastructure, poverty, access, Asia, Jakarta

* Abidin Kusno, Bharat Punjabi, Michael Leaf, Sylvy

Jaglin, Olivier Coutard, Steve Graham, and three anon-

ymous reviewers provided helpful comments on various

drafts of related material. Research assistance from Er-

nst Jan Martijn and Nur Endah Shofiani is gratefully

acknowledged, as is the role played by host Professor

Teti Argo at ITB during research visits to Indonesia. The

research upon which this paper is based was funded by a

grant from the Social Sciences and Humanities Research

Council of Canada. Much of the material is drawn from

a background report originally commissioned by the

UNDP HDRO as part of the 2006 Human Development

Report. Assistance from archival staff at KIT (Amster-

dam) and KITLV (Leiden) during research in 2003 is

gratefully acknowledged, as is that of water supply

managers at PAM Jaya and staff at numerous govern-

ment departments and NGOs during fieldwork con-

ducted in Jakarta in 2001, 2004, and 2005. The usual

disclaimers apply. Final revision accepted: September 8,2007.

1. INTRODUCTION

Inequitable access to water supply and sanita-tion has been characterized as a critical develop-ment challenge for the South. Halving by 2015the proportion of people without sustainableaccess to safe drinking water and basic sanita-tion is one of the Millennium DevelopmentGoals (UNDP, 2003). The World Health Orga-nization estimates that 1.1 billion people world-wide do not have access to safe drinking water(WHO/UNICEF, 2004; WHO, 2000). Anincreasing proportion of users without accessto adequate water supplies live in urban areas;poor families in large cities in the South fre-quently do not have networked water supply ac-cess (UNCHS, 2003, 2006; UNDP, 2006;UNWWAP, 2003, 2006). The most recentassessment by UNCHS estimates that 970 mil-lion urban dwellers are without access to ‘‘ade-quate’’ water supply (cumulative total forAfrica, Asia, and Latin America and the Carib-bean) (UNCHS, 2006). Given the lack of reli-able data, 1 current estimates of urbandwellers without access to ‘‘adequate’’ provision

189

for water supplies are necessarily highly uncer-tain, but it seems likely that the true numberof urban dwellers with inadequate provision

1

1892 WORLD DEVELOPMENT

for water supply is significantly underestimatedby governments and international agencies, andis growing as rapid urbanization continues inmany regions (UNCHS, 2003, 2006).

This paper examines factors that explain thepersistent failure of both public and privatewater supply system operators to achieve highrates of individual network connections to poorhouseholds in urban areas. 2 Section 1 presentsa conceptual framework of ‘‘governance fail-ure,’’ which focuses on the institutional dimen-sions of urban water supply provision. Indeveloping the conceptual framework in Sec-tion 1, the paper summarizes the recent litera-ture pertaining to the ‘‘ownership’’ versus‘‘institutions’’ debate, focusing on the putativemerits of public versus private management ofwater supply systems. The analysis concludesthat institutions have an important impact onmanagement and performance, and argues thatmore systematic attention should be given tothe institutional dimensions of water supplyprovision with respect to urban poor house-holds. Building on this discussion, we presenta concept of ‘‘governance failure’’ as it appliesto urban water supply to poor households, pos-iting that governance failures apply not only towater supply providers and governments, butalso to poor households, whose capability toconnect may be undermined by a range of eco-nomic and non-economic factors.

In subsequent sections of the paper, this con-ceptual framework is applied to the case of Ja-karta, using primary data from householdsurveys, archives, water utility reports, GIS-based mapping, and interviews with water man-agers, government officials, lenders, andNGOs. 3 Section 2 documents the differentia-tion of access to water supply in Jakarta, andargues that the spatial dimensions of networkaccess and exclusion—that is, the urbanmacro-geography of network distribution—are an important and under-documented aspectof the failure to achieve high connection ratesfor poor households. Sections 3 and 4 of the pa-per attempt to explain this differentiation of ac-cess through presenting an analysis ofgovernance failures, respectively, pertaining toJakarta’s water supply utility (under both pub-lic and private management) and poor house-holds. Specifically, Section 3 analyzesgovernance failures pertaining to the water sup-ply utility and municipal government, which in-clude water supply utility governance norms,land use policies and related municipal deci-sion-making processes, the water supply utility

business model, and discriminatory connectionpolicies on the part of water utilities linked totariff-related economic disincentives for con-necting poor households. Section 4 then ana-lyzes sources of governance failure pertainingto the capability of individual households toconnect to the water supply system, which in-clude connection fee policies, transaction costs,housing and residency status (tenure), securityof water supply, and perceptions of water qual-ity. The findings indicate that governance fail-ures produce important disincentives for thewater supply utility to connect poor householdsand for poor households to choose to connectto the water supply system.

The final section of the paper argues that theinstitutional and spatial dimensions of watersupply are under-documented aspects of thefailure to supply water access to poor house-holds in urban areas, and concludes with a dis-cussion of policy implications.

2. MARKET FAILURE, STATE FAILURE,GOVERNANCE FAILURE: EXPLAINING

LACK OF ACCESS TO NETWORKEDWATER SUPPLY BY THE URBAN POOR

Supplying water to the world’s poor has beenhigh on the agenda of the international commu-nity for decades, yet lack of universal access tourban water supply has persisted despite sus-tained and significant investment by bilateralaid agencies and multilateral financial organiza-tions. Water supply figured prominently on theagendas of the Stockholm Environment (1972)and Vancouver Habitat (1976) conferences. Atthe UN’s Mar del Plata conference (1977), theUnited Nations Water and Sanitation Decadewas formally agreed; over the period 1981–90,bilateral aid and multilateral finance were direc-ted toward water supply projects in the South inunprecedented amounts (WHO, 1992). At theend of the decade, more people (in absoluteterms) were being supplied with ‘‘improvedwater supplies’’ than ever before; during thedecade, it was estimated that 1.2 billion peoplegained access, for the first time, to a safe andadequate water supply (WHO, 1992). Yet, inmany countries, the increase in supply hadfailed to keep pace with the population growth,and growing numbers of people remained with-out access to the World Health Organizationminimum of 25 L per person per day of potablewater; the number without access to a safe andadequate water supply fell by only 450 million

GOVERNANCE FAILURE 1893

during the same period (WHO, 1992). Muchattention during these decades was devoted towater supply in rural areas, but more recentlythe particular problems posed by the lack of ac-cess to water supply by poor households in ur-ban areas in the context of rapid urbanizationhave been given greater attention (UNCHS,2003, 2006; UNDP, 2006; UNWWAP, 2003,2006; WHO, 1992).

The failure to achieve improved rates of con-nection of poor households to water supply,particularly in urban areas, was part of the rea-son for the emergence over the past decade of awide-ranging debate on new approaches towater supply in the South. An importantdimension of this debate has been the questionof the relative importance of ownership versusinstitutions 4 in influencing the performanceof water supply utilities, and network utilitiesmore generally (e.g., Newbery, 2000). Thisquestion has particular salience given the signif-icant increase in private sector participation(PSP) in water supply in urban areas in devel-oping countries over the past two decades,and the strong support for some bilateral andmultilateral agencies for private sector involve-ment. Indeed, much recent research focusing onthe institutions/ownership debate with respectto water supply focuses on the question of therelative merits of the public and private sectors,in the context of heated debate about the roleof ‘‘state failure’’ and ‘‘market failure’’ in lowpenetration rates for water supply utilities.

A significant emphasis of the recent debateover the role of ownership versus institutionshas focused on the question of the relative mer-its of the public and private sectors in manag-ing water supply systems. What does thisresearch tell us? Reviews of the performanceof water supply utilities in Asia, the UnitedStates, the United Kingdom, and Europe indi-cate that ownership (public or private) doesnot predict the efficiency of water service pro-viders (Bayliss, 2003; Braadbaart, 2002;Estache & Rossi, 2002; Hodge, 2000; Hunt &Lynk, 1995; Kirkpatrick, Parker, & Zhang,2006; Lobina & Hall, 2000; Prasad, 2006;Renzetti & Dupont, 2004; Wallsten & Kosec,2005). More generally, the public administra-tion literature states that there is no associationbetween the formal-legal status of a public ser-vice provider and its service delivery perfor-mance (see Holmes, 2000, with respect towater utilities and (Verhoest, Peters, Bouckaert,& Verschuere, 2004) for a more general argu-ment). According to this view, institutions 5

have a greater impact on water utility perfor-mance than ownership (Budds & McGranahan,2003; Gleick, Wolff, Chalecki, & Reyes, 2002;Martin, 2004).

However, other studies have found that own-ership does have an effect. Proponents of PSPsargue that private sector involvement improvesperformance through, for example, higher effi-ciency and cost recovery, enabling additionalsources of finance, or higher connection ratesfor poor households (Cross & Morel, 2005;Johnstone & Wood, 2001; Nickson & Franceys,2003; Shirley, 2002; Winpenny, 1994; WorldBank, 1994, 1997a, 2004a). Conversely, oppo-nents of PSPs have argued that private sectorparticipation negatively affects performancethrough raising the cost of capital, reducinglong-term investment in infrastructure repairand replacement, increasing corruption, orreducing affordability due to tariff increases(Bayliss, 2002; Bayliss & Fine, 2007; Bond,2002, 2004; Hukka & Katko, 2003; McDonald& Ruiters, 2005; Wilder & Romero Lankao,2006). In short, both proponents and oppo-nents of private sector involvement attributean important role to the ownership of watersupply system managers.

To some degree, these divergent interpreta-tions of the relative importance of ownershipversus institutions are dependent upon differentdefinitions of ‘‘performance’’: for example,studies focusing on efficiency have found littleeffect of ownership, whereas studies focusingon distributional concerns (e.g., tariff rates,connection rates to poor households) havefound a greater effect (Prasad, 2006). Thesedivergent interpretations may also be due toideological bias: opponents (or proponents) ofprivatization may be more likely to focus uponnegative (or positive) effects of involving theprivate sector, thereby emphasizing the impor-tance of ownership. Notwithstanding their dif-fering conclusions, most recent research onwater supply utilities acknowledges that institu-tions do matter. Inadequate or absent regula-tion, for example, is widely acknowledged tocontribute to poor PSP outcomes (ADB,2003a; Anwar, 2001; Brown, 2001; Budds &McGranahan, 2003; Franceys & Weitz, 2003;Gutierrez, Calaguas, Green, & Roaf, 2002;Robbins, 2003; UNDP, 2003). In short, the lit-erature suggests that whereas institutions aredeterminant of utility performance, the effectof ownership (positive, negative, or neutral) isdisputed. In reality, of course, it may be thecase that ownership and institutions are to


some degree inter-related; for example, a changein ownership will result in (and perhaps be en-abled by) a change in institutions (Ostrom,1990; Ostrom, Schroeder, & Wynne, 1993).

Despite the acknowledged importance ofinstitutions in the literature, institutional analy-sis has not been prioritized in the recent debateon urban water supply, particularly with re-spect to PSPs (although, for more general dis-cussion, see Saleth & Dinar, 2000, 2005;Trawick, 2003). Much of the literature seeksto assess the merits of public and/or privateproviders, relying (at times implicitly) on con-cepts of market failure and state failure. Theconcepts of market and state failure are admit-tedly important, particularly when consideringthe case of water supply in cities in the South,where water supply networks are not universaland where non-networked sources include gov-ernment, NGO, and private sector providersdrawing on different sources of water (groundwater, surface water, rainwater), and employ-ing different technologies of water provision(networks, wells, treated bottled waters, bulkwater deliveries, and rainwater harvesting)(Collignon & Vezina, 2000; UNCHS, 2003,2006). This is the case because networked watersupply in many cities is not a natural monop-oly. Rather, in many cases, state and privatesector actors 6 are simultaneously involved invarious aspects of water supply, and often com-pete for clients. Both market failure and statefailure may thus characterize specific aspectsof urban water supply delivery.

However, insofar as the concepts of marketand state failure tend to encourage a focus onownership, rather than on institutions, thecomplex set of institutional factors underpin-ning the failure of both public and private net-worked water supply operators to effectivelyand universally supply poor households maybe insufficiently understood. Recent debateson water governance, in contrast, focus explic-itly on these institutional factors, where watergovernance is defined as the range of political,organizational, and administrative processesthrough which stakeholders (including citizensand interest groups) articulate their interests,exercise their legal rights, take decisions, meettheir obligations, and mediate their differences(adapted from Rogers & Hall, 2003 andUNDP, 2007; see also Nunan & Satterthwaite,2001). Although water supply governance istypically defined as a process dominated bygovernments, water supply utilities, and com-panies, the literature on ‘‘distributed gover-

nance’’ reminds us that it is also a process inwhich non-state actors—including citizens—can participate, although their degree of partic-ipation may vary significantly (Pierre, 2000;Pierre & Peters, 2000; Strange, 1996; Rhodes,1996).

Following from this definition, ‘‘governancefailure’’ occurs when institutional dimensionsof water management and decision-making donot effectively take into account the needs ofpoor households, creating disincentives for thewater supply utility to connect poor householdsand/or for poor households to connect to thenetwork (Table 1). The concept of governancefailure thus requires analysis of the bases andprocesses for decision-making across the fourkey dimensions of water management: adminis-tration, delivery (technical services), financialand economic management, and political over-sight (UNCHS, 2003). This formulation of theconcept of governance failure thus suggests thatdecision-making structures and related institu-tions may contain systematic biases againstpoor households despite, for example, officiallystated pro-poor policies, and independent ofthe ownership status (public or private) of thewater supply network and its manager. This as-pect of governance failure is explored in greaterdetail in Section 3.

This definition of governance failure hasobvious overlaps with Sen’s concept of capabil-ities (Robeyns, 2005; Sen, 1992, 1999). Follow-ing Sen, the term capability is defined asindividual’s ability to achieve desired function-ings, where functionings represent actualachievements in ‘‘doing or being’’—both tangi-ble and intangible (e.g., being well-nourished,adequately hydrated, healthy, well educated,or having self-respect due to active participa-tion in community decision-making). A ‘‘capa-bility set’’ is defined as the set of attainablefunctionings that an individual is able toachieve (Sen, 1992, 1999). The concept of acapability set is analytically useful because itprovides a means of explaining why differentindividuals are able to differentially mobilizespecific commodities to achieve certain functi-onings. This approach is also useful because itemphasizes agency, insofar as individualchoice, values and preferences play an impor-tant role in determining which functioningsare chosen from a possible capability set(Ibrahim, 2006). The implication follows thatthe ‘‘capability set’’ of having, for example, ahigh level of water-related hygiene and healthmay be met through a variety of functionings,

Table 1. Connecting poor households, to networked water supply: selected examples of market, state, and governancefailures

State failure Water supply systems owned and operated or regulated by governments may fail to

operate effectively when one or more of the following occur:

1. Rent-seeking (by officials)2. Unincorporated externalitie3. Poacher–gamekeeper problem (if both supplier and regulator are public)4. Regulatory capture (if supplier is private)

Market failure Water supply networks operated by private companies may fail to operate efficiently

when one or more of the following occur:

1. Imperfect competition2. Asymmetric information (between regulator and company)3. Unincorporated externalities4. Public good (health benefits of universal water supply provision are non-exclud-able and non-rivalrous)

Governance failure The decision-making process for water management may fail to address the needs of

poor households because of

1. Absence of consumer entitlements to basic services (e.g., lack of universal servicerequirement on the part of utility)2. Political disenfranchisement (e.g., lack of ‘‘voice’’ on the part of poor households)3. Culture of governance (e.g., elite-focused, top-down)4. Economic disincentives for connecting poor households

Individual households may be subject to institutions, incentives, or other factors, which

undermine their capability to connect to the water supply system

1. Tenure system (lack of clear property rights)2. Lack of skills (e.g., literacy) facilitating interaction with service provider3. Cultural beliefs (e.g., appropriate water treatment protocols)4. Tariff structure (e.g., high connection fees)

Source: Authors.


not all of which correspond to the conventionalview of the optimal approach to water supplydelivery (i.e., individual household networkconnections). In other words, adopting a capa-bility approach opens up two possibilities: thatexogenous factors constrain the capability setof individuals (precluding their choice to con-nect to the water supply network) and thatpoor households may be making a rationalchoice not to connect to the water supply net-work despite its inclusion in their capabilityset. These aspects of governance failure are ex-plored in greater detail in Section 4.

In summary, the concept of governance fail-ure may usefully be applied to the study of watersupply to poor households in urban areas in theSouth, because it (1) formalises an analysis ofinstitutions and associated incentives to whichgovernments, water supply utilities, and con-sumers are subject; and (2) enables the analysisof decision-making processes and related insti-tutional incentives on the part of both water

supply providers and consumers. Moreover,and as argued in the following sections of thepaper, this conceptual framework provides in-sight into the failure of both public and privatewater supply system operators to expand watersupply access to poor households. Specifically, afocus on governance failure allows identifica-tion of institutional barriers common to boththe public and private sectors (which suggeststhat the debate over the relative merits of publicversus private ownership and/or managementof water supply systems has, to some degree,missed the point (Budds & McGranahan,2003). To illustrate these arguments, Sections3 and 4 of the paper apply this conceptualframework to the case of Jakarta, respectively,providing empirical examples of governancefailures pertaining to the water supply utilityand the poor households. Prior to this, Section2 briefly provides background on the case study,providing evidence on the degree of differentia-tion of access to water supply in Jakarta, which


emphasizes the degree to which spatial and so-cio-economic exclusion from networked watersupply access overlap within the city.

3. THE GEOGRAPHY OF EXCLUSION:DIFFERENTIATED ACCESS TO WATER

SUPPLY IN JAKARTA

(a) Access to water supply in Jakarta

Several studies have characterized Jakarta’swater and sanitation sector as one of the weakestin Asia (Brennan & Richardson, 1989;Leitmann, 1995; McGranahan, Jacobi, Song-sore, Surjadi, & Kjellen, 2001; McIntosh, 2003;World Bank, 2004b). Official estimates of theproportion of the city’s population with watersupply network connections in the home(‘‘household connections’’) range from 46% to56% (BPS, 2005; Jakarta Water Supply Regula-tory Body, 2004). 7 Unofficial estimates, whichattempt to account for the large number of infor-mal residents in the city, suggest that only 25% ofDKI Jakarta’s true population is being served(Tutuko, 2005). Domestic water consumptionis estimated to be between 70 and 80 L per capitaper day—one of the lowest of the 18 large Asiancities surveyed by the ADB in 2002 (ADB,2003b; McIntosh, 2003).

The correlation between poverty and lack ofaccess to a household connection has been doc-umented in household surveys of Jakarta(Crane, 1994; Forkami/RTI, 2002, 2003;McGranahan et al., 2001; Shofiani, 2003). Thiscorresponds with national data which indicatethat the urban poor lack access to water supplyacross Indonesia: data from Indonesia’s censusindicate that only 16% of the urban ‘‘verypoor’’ (with monthly incomes of Rupiah800,000 (approximately $80 USD)) have house-hold connections, while 36% of the urban pop-ulation is connected, on average, across thecountry (Woodcock, 2005). Those not con-nected to Jakarta’s municipal water supply sys-tem rely on a variety of sources (shallow anddeep groundwater; surface water via the muni-cipal water supply network; spring water; and(less frequently) direct use of rainwater, rivers,and streams) and distribution methods (house-hold wells or rainwater collection systems;water vendors; bottled water; public stand-pipes; small, private networks connected todeep wells; and water trucks) (Berry, 1982;Gilbert & James, 1994; Lovei & Whittington,1993; McGranahan et al., 2001). As with other

large cities in developing countries, many ofthese water supply methods are more expen-sive, per unit volume, than piped water supply.Thus, given their lower incomes, many poorhouseholds pay a much higher proportion oftheir income for water than wealthierhouseholds do. Wealthier households with anetworked connection, in other words, receivewater at a lower cost per unit volume,spending lower proportions of income formuch greater quantities of water, in a patterntypical of many cities in developing countries(see, e.g., Cairncross, Hardoy, & Satterthwaite,1990; Gulyani, Talukdar, & Kariuki, 2005;Swyngedouw, 1997).

Most of the above-mentioned water sourcesare characterized by poor water quality. Jakar-ta’s small municipal sewerage system connectsless than 2% of households, a legacy of govern-ment policy treating sewage as a ‘‘private con-cern’’ (ADB, 2003b; Argo, 1999; Cowherd,2002; World Bank, 1993). The vast majorityof wastewater is disposed directly to rivers, ca-nals, or to (often poorly functioning) septictanks (Crane, Daniere, & Harwood, 1997;McIntosh, 2003; Surjadi, 2002). The resultingcontamination of surface water sources by sew-age and industrial effluent is exacerbated by thelack of an effective waste collection system inthe city, with household waste collecting in ca-nals which provided water sources and flooddrainage for the city (see, e.g., Porter, 1996).Open canals, largely conduits for sewage, regu-larly overflow into city streets during the rainyseason. As a consequence, poor water quality inthe piped network (partially reliant on surfacewater sources within the city), and in shallowgroundwater (the source for the majority ofthe city’s poor residents) is of particularconcern. Contamination by wastewater andindustrial effluent, as well as salinization—purportedly due to seawater infiltration due toover-pumping—has polluted Jakarta’s shallowaquifer, the sole household source of supplyfor many poorer families in many areas of thecity (Braadbaart & Braadbaart, 1997). Riversand canals are usually too polluted to use evenfor washing clothing. Nor is the water deliveredthrough the network potable, due to the poorquality of the raw water available to the watersupply utility and the poor infrastructural qual-ity of the distribution network, leading to infil-tration and contamination in the distributionmains after water has left the treatment plant.Indeed, public health studies repeatedly find fe-cal coliform contamination in Jakarta’s net-


worked water supply system, and residents areadvised to boil their water. The public healthimpacts of this situation are predictable, andhave been well-documented: high rates ofwater-related diseases, including gastrointesti-nal illness due to contaminated water and par-asite-related illnesses due to poor drainage,particularly in poorer areas (Agtini et al.,2005; Leitmann, 1995; McGranahan et al.,2001; Simanjuntak et al., 2001; Surjadi, 2003).Results of the first community-based surveil-lance study of diarrhea in Jakarta found thatdiarrheal diseases, many of them water-related,are the third leading cause of morbidity and theleading cause of morbidity in infants (Agtiniet al., 2005).

(b) Jakarta’s water network as elite‘‘archipelago’’

Although available data indicate that a sig-nificant proportion of Jakarta’s residents arenot connected to the water supply network,no study has been conducted to date document-ing the spatial dimensions of access. Where areJakarta’s residents connected (or not) to thewater supply network? This section answersthis question through presenting a reconstruc-tion of the growth of the city’s network, incor-porating primary archival interview data usingGIS-based mapping. The analysis is necessarilyhistorical, drawing on colonial and post-colo-nial archival data, because water infrastructurenetworks have one of the longest turnovertimes of any dedicated utility infrastructure;in many cities around the world (including Ja-karta), it is common to find pipes over 100years old still in service. Given this longevity,water networks physically embody successivephases of management and investment; histori-cal choices thus shape contemporary con-straints on water supply network managers.

Our archival research on the original colonialnetwork indicates that Jakarta’s original watersupply system was deliberately limited to the‘‘European’’ urban population. A subsequentspring water-fed network built in 1920 had onlylimited extension into the ‘‘kampungs,’’ 8 basedon a dual design standard: 140 L/hhold/day tobe distributed to 90% of the European house-holds in Batavia, in contrast to an anticipateddelivery of only 65 L/hhold/day to only 33%of the native population (van Breen, 1916;Van Leeuwen, 1917). Penetration of the watersupply network into non-European neighbor-hoods remained limited. The result was a differ-

entiation of access and consumption levels: in1929, the European population, comprisingonly 7% of the population, consumed 78% ofthe volume of water supplied to residential cus-tomers (Eggink, 1930; Maronier, 1929). 9

In the decades following Independence, watersupply in Jakarta continued to be highly differ-entiated, and was characterized by many simi-larities to colonial patterns of water provision.In part, this was because of technical con-straints: surface water supply continued to bedistributed through the colonial piped network,and was thereby restricted to more affluentareas of the city. But this is only a partial expla-nation, as major investments were made in thefirst decade following Independence to bothrehabilitate and expand the water supply net-work. In particular, two major water treatmentplants were completed in the 1950s and 1960s,adding 3000 L/s to Jakarta’s water productioncapacity. Networks to distribute this water werenot, however, extended universally across thecity, despite the fact that production capacityexceeded distribution capacity in the decadesfollowing independence (Martijn, 2005). Socialpolicy choices made by the government are anequally important explanatory factor: in re-sponse to rapid (and usually technically illegal,given the lack of residency permits) in-migra-tion of refugees, and then immigrants to Jakartafollowing Independence, the provision of gov-ernment services in non-official neighborhoods(without tenure status) was curtailed or prohib-ited. Moreover, the government’s emphasis onconverting Jakarta into an international show-piece to symbolize Indonesian unity and devel-opment meant that financial resources weredirected almost solely to showcase projects(Kusno, 1997, 2000). Network expansion wasdeliberately limited to upper class residentialareas (such as Keborayan Baru), with new ser-vice mains following Jakarta’s new ‘‘modernstrip’’ of elevated highways and flyovers, Indo-nesia’s first international standard hotel (HotelIndonesia), the highrise developments alongthe main thoroughfare Jalan Thamrin-Sudir-man, and the Asian Games Complex, therebyexcluding the ‘‘unmodern’’ spaces and popula-tions thought to ‘‘lower the status of the nation’’(Abeyasekere, 1989; Kusno, 2000; MacDonald,1995), from both spatial proximity to, and ser-vices from, the network (Map 1).

Map 1 presents the results of a GIS-basedmapping exercise correlating network distribu-tion with land use. The analysis demonstratesthat the distribution of reticulation (tertiary)

Map 1. Type of income and type of water supply. Source: Compiled by the authors (base maps supplied courtesy of Pam

Jaya, household census data from BPS—Indonesian Office of Statistics).


networks is concentrated in key economiczones (the Tanjung Priok port area, the Pulog-adung Industrial Estate, and the commercial

district surrounding the Hotel Indonesia) andelite residential zones (Tebet, Menteng, Ke-mang, Pondok Indah, and Keborayan Baru).


Most poor neighborhoods (still colloquiallyknown as ‘‘kampungs’’) continue to be ex-cluded from network access due to the spatialdistribution of water mains and reticulation(tertiary service) pipes, as indicated by theleft-hand inset, indicating that lower or no cov-erage is associated with higher rates of poverty(using percentage of slum dwellers per districtas a proxy). Moreover, areas of low or no pres-sure, as indicated on the map, tend to be con-centrated in the poorest area of the city—North Jakarta—the area of original colonialsettlement and of the oldest, and most poorlymaintained infrastructure. Even in neighbor-hoods with water mains, access to networksvaries at an extremely fine spatial resolution,often on a street-by-street basis (as illustratedby the right-hand inset on Map 1): reticulationnetworks are densely and evenly distributed inhigh- and middle-income streets, but much lessso in low-income streets: reticulation networksare densely and evenly distributed in high-and middle-income streets, but much less soin low-income streets. This pattern was estab-lished during public management of the watersupply network, and remained largely stablefollowing the introduction of private sectormanagement of the water supply system, forreasons explored below.

In summary, the above analysis documentsthe link between inequitable water supply ac-cess, socio-economic status, and spatial distri-bution of water supply infrastructure inJakarta. These factors—particularly the spatialdimension—are important barriers to watersupply access by poor households (and are of-ten only anecdotally acknowledged, rather thansystematically documented). This suggests thatthe term ‘‘network’’ is a misnomer when ap-plied to Jakarta’s water supply system, whichis more properly described as an ‘‘archipelago,’’largely confined to elite residential and indus-trial areas of the city (Bakker, 2003).

4. GOVERNANCE FAILURE AND THESUPPLIER: WHY WATER UTILITIESCHOOSE NOT TO CONNECT POOR

HOUSEHOLDS

The previous section documented how the pat-tern of highly differentiated access emerged;here, we deploy the concept of governance fail-ure to explain why this has occurred. Specifically,this section documents governance failures per-taining to the water supply utility (under both

public and private sector management) andmunicipality: failures of the water managementdecision-making process and associated institu-tions to account for the needs of poor house-holds. Our evidence is drawn from archivalresearch, as well as from interview data collectedin 2001, 2005, and 2006. 10 Prior to exploringspecific examples of governance failure, it shouldbe emphasized that the spatial heterogeneity ofthe network as presented in Section 2 was not so-lely due to a lack of water supply or water re-sources availability during the colonial andmuch of the post-Independence period. Indeed,the colonial water supply system could theoreti-cally have satisfied Jakarta’s entire population(of approximately 800,000) up until the end ofthe 1940s, assuming a per capita water demandof 50/person/day, respectively, and reasonable(e.g., 30%) distribution losses (Martijn, 2005).This surplus of potential over actual supply per-sisted post-independence, when productioncapacity consistently outstripped distributioncapacity from the 1960s onwards (Martijn,2005). In particular, from 1965 to the late-1980s, the rehabilitation of existing infrastruc-ture, and the construction of an additionallarge-scale water treatment plant increasedwater supply production capacity threefold,but by 1990 the water supply system still onlydelivered 40% of the potential volumes of treatedwater (JICA, 1997), and the provision of pipedwater supply still only extended to less thanone-quarter of the city’s population (Porter,1996). Yet, despite excess production capacityand a rapidly growing population, 11 the city’swater utility embarked only on limited exten-sions of the distribution network, 12 and in-creased production capacity largely in supportof industrial users in the 1970s and 1980s (Salim,2005; Tutuko, 2005). In other words, a shortageof water resources was not the primary, or thesole reason for the failure to extend water supplynetwork connections to poor households. Scar-city of potable water was indeed experienced inthe city, but this was ‘‘second-order’’ or ‘‘socialresource scarcity’’ rather than ‘‘first-order’’ or‘‘natural scarcity’’ (cf. Ohlsson, 2000).

What, then, explains the failure of PAM Jaya,the city’s water supply utility, to extend net-works to poor households? A primary factoridentified by interviewees is the culture of gover-nance within urban government in Indonesia,which does not prioritize the poor. The waysin which this low priority has been expressedhave changed over time. Post-independence,urban services provision and the urban


environment were given relatively low priority(Chifos & Hendropranoto, 2000; Firman &Dharmapatni, 1994; Ford, 1993; World Bank,2004b), despite sporadic national government-led development plans to accelerate servicedelivery (e.g., Silas, 1992; World Bank, 1999).Water treatment and distribution, as ‘‘invisible’’infrastructure, was a relatively low priority forthe newly independent government dedicatedto modernizing services for the elite, rather thanuniversalizing services for the masses (Kusno,1997; Leclerc, 1993). This provides a partialexplanation for the reported lack of official con-cern about the fact that the majority of the ur-ban population could not afford to connect tothe network (Fischer, 1959; Kusno, 2000).

Subsequently, low priority was placed on theprovision of piped water; this was rationalizedas a policy to discourage rural migrants, whowere blamed for over-taxing the city’s publicservices (KIP, 1976). However, anti-migrationpolicies are not the sole factor; coordinationwith PAM Jaya to extend piped water supplyinto poorer neighborhoods did not greatly im-prove after the ‘‘closed city’’ policy was relaxedin 1976 (Taylor, 1983). Indeed, the first projectfocused on urban water supply to the poor oc-curred only in the 1990s (the World Bank-fi-nanced Pam Jaya Supply ImprovementProject, 1990–97); although it resulted in over200,000 new connections, the expansion of pro-duction and distribution capacity did not catchup to increasing (potential) demand due to pop-ulation growth (Cestti, Batia, & Van der Berg,1994; PAM JAYA, 1992a, 1992b; Porter,1996; JICA, 1997; World Bank, 1998). By theend of the 1990s, only approximately 10% ofkampung residents had household water con-nections (Azdan, 2001); the remainder wereserved with public hydrants or water vendors,consuming only 7% of the water distributedby PAM Jaya, versus the 57% consumed byhouseholds with individual household connec-tions (JICA, 1997).

A second source of governance failure identi-fied by interviewees arises from land-use poli-cies, and related decision-making processes atthe municipal level. Official development plansfor Jakarta encouraged an east-west patternof urban development, attempting to avoidexpansion into irrigated agricultural areasnorth and south of the city (JICA, 1997). Yet,despite the failure of planning controls to stemurban sprawl, PAM Jaya’s network expansionhas been limited to target zones in official plan-ning, thereby excluding zones of the city where

active development was occurring. This hasbeen exacerbated by the lack of participatorygovernance mechanisms on the part of PAMJaya that would allow the concerns of poorhouseholds to ‘‘filter up’’ to utility managersor political leaders, who may be sometimesunaware of the numbers of unconnected resi-dents, or the cost to the unconnected poor ofobtaining water from alternative sources(Woodcock, 2005).

A third source of governance failure is thebusiness model adopted by Indonesian watersupply utilities (Woodcock, 2005). Water utili-ties in Indonesia are controlled by local govern-ment; senior appointments may in someinstances be guided by political patronagerather than technical requirements. Employ-ment in a water utility in Indonesia has conven-tionally been associated with low social status(Martijn, 2005) and low educational levels(World Bank, 1997b); the associated impactson morale and technical expertise of staff haveundermined the efficiency and productivity ofmost water utilities in Indonesia. Moreover, lo-cal governments have typically been unwillingor unable to make politically unpopular deci-sions (such as raising tariffs) or require waterutilities to improve performance (e.g., throughmeasurable performance targets). The treat-ment of water utilities as ‘‘cash cows’’ (notablythrough the payment of annual dividends to themunicipality) 13 has in some cases distortedlong-term planning processes, reducing invest-ment in infrastructure maintenance and renew-al (World Bank, 1997b). The absence ofrepresentatives of user groups (e.g., residentialand business users) on the Board of Supervisorsof most water utilities in Jakarta has furthercontributed to the lack of counterweight topolitical influence on water supply utility deci-sions (World Bank, 1997b).

A fourth source of governance failure, relatedto the choice of business model, has been theview, often implicit in PAM Jaya policy, that‘‘kampung’’ neighborhoods are literally ‘‘off-limits’’ because they are ungovernable andunserviceable. This view was reiterated in inter-views with utility staff, who alleged that in someneighborhoods local mafia groups control pub-lic standpipes, actively discouraging (sometimesviolently) attempts to install distribution net-works or substitute household water connec-tions for water hydrants (Crane, 1994;Pandjaitan, 2004). Others referred to the diffi-culties created due to the low social capital ofpoor households; staff may be reluctant to deal


with the poor because, for example, of illiteracyor their lack of ability to correctly fill outforms. These views were both reflected andreinforced by PAM Jaya policies which focusedon ‘‘serviceable’’ areas, and required full costrecovery from customers after the early-1970s(PAM JAYA, 1992b; Taylor, 1983). 14 As aconsequence, PAM Jaya preferred to limit theextension of distribution networks to neighbor-hoods where the ‘‘user-fee’’ schemes used to fi-nance network extension meant that costscould reliably be recovered. 15 The discrimina-tory social policy of PAM Jaya during this per-iod is reflected in its own admission that it was‘‘best situated to serve well established, formalareas comprising concentrated groups of users,rather than newly developed and widely scat-tered areas [i.e., kampungs].’’ (Argo, 1999, p.71). The limited ability of PAM Jaya to extendwater supply network into poor areas is corrob-orated in the latest Master Plan for Jakarta’swater supply, which records the absence of apoverty reduction strategy in network exten-sion policy until the late-1980s (JICA, 1997)and notes the absence of policy targets for sup-ply coverage (in % terms of total DKI popula-tion) until the 1970s (World Bank, 1991).Indeed, PAM Jaya’s reported coverage figuresfor the city were usually stated as proportionsof the total population in the target ‘‘servedarea,’’ excluding kampungs altogether.

In addition, tariff-related incentives may playan important role in governance failure. Forexample, the rising block tariff structure initiatedin the 1980s and maintained under both publicand private management 16 created (and stillcreates) a strong disincentive for the watersupply utility to connect the poor. Public hy-drants—usually built in kampungs—are char-ged higher volumetric tariffs than individualhouseholds, creating a counterintuitive cross-subsidy from poor to middle and upper class cus-tomers. 17 Banded tariff structures with a risingblock tariff beginning with rates below produc-tion cost create a disincentive for providing di-rect network connections to poor customers,who pay lower amounts per unit volume; largenumbers of poor customers thus threaten to de-crease water company revenues, and could theo-retically result in revenue per unit volume fallingbelow marginal cost (Whittington, 1992). Evenin areas with the possibility of network connec-tions, higher tariffs (with the volumetric tariffdoubling in less than a decade in real terms),higher connection charges, and a deposit feemeant a high initial fixed cost for a household

water supply connection, prohibitive to poorhouseholds with low and often fluctuating in-comes. As a result, only a small proportion ofthe new connections made in the 1990s werefor very poor households (Cestti et al., 1994;JICA, 1997; Porter, 1996).

Tariff-related disincentives continued to affectnetwork management following the private sec-tor contract initiated with two consortia (divid-ing the city in half, with subsidiaries of Ondeo/Lyonnaise des Eaux and Thames Water runningthe network in the western and eastern halves ofthe city, respectively) in 1998 (Argo & Firman,2001; Bakker, 2007). Under the terms of the ini-tial contract, each consortium was to receive afee on the basis of the volume of water suppliedand billed. With no direct equity stake, and withprofit de-linked from cost recovery rates, theinternational water companies thus sought tominimise the risk inherent in cost recovery. Anadditional safeguard was built into the paymentmechanism: an indexation formula, linked to theRupiah-US dollar exchange rate and the (Indo-nesian) inflation rate was built into the ‘‘watercharge’’ formula used to determine paymentsmade to the private operators. Cost recoveryand currency risks, in other words, were to beborne by the local government. However, rapidcurrency devaluation in 1998 resulted in a signif-icant decrease in receipts in dollar terms. Giventhe political unrest in Jakarta and resistance bywater sector unions, the local government didnot implement agreed-upon tariff increases(Harsono, 2005). The gap between the watercharge required for compensating the privatecompanies and the average water tariff increaseddramatically (Figure 1). Prolonged negotiations(lasting several years) between the concession-aires and the city resulted in eventual tariff in-creases, but these did not raise the tariff abovethe water charge until early 2004 (Jakarta WaterSupply Regulatory Body, 2004). PAM Jaya (andthus the local government) bore the sole risk forthe revenue shortfall, and became increasinglyindebted to the two private concessionaires. 18

This indebtedness reinforced disincentives toconnect loss-making poor households; unsur-prisingly, the new connections made by the pri-vate concessionaires focused preferentially onmiddle-class customers, which received 58% ofnew connections between 1998 and 2004, withlow income and ‘‘very poor’’ households receiv-ing only 24% of new connections (Table 2).‘‘Pro-poor’’ activities on the part of the privateconcessionaires were constrained by low willing-ness-to-pay for connection fees on the part of

Figure 1. Water charge and water tariff. Source: Jakarta

Water Supply Regulation Body, Personal communication

with authors.


poor households, and were limited to a fewshowcase projects such as that in the neighbor-hood of Marunda (Bakker, 2007; BPD, 2003).

5. GOVERNANCE FAILURE AND THECONSUMER: WHY POOR HOUSEHOLDS

CHOOSE NOT TO CONNECT

In this section of the paper, we document gov-ernance failures affecting poor households. Ourevidence is drawn from primary survey databased on a survey of 110 households in six Jakar-

Table 2. Tariffs and

Averagetariffs pertariff bandRp./M3(2005)

Monthly fixedcharges (Rp.) (2005)

Taridescriptio

I 550 4,695 Social institutifacilities) and

IIa 550 5,060 Public hospitahou

IIb 2,450 10,440 Low-incomIIIa 3.500 11,950 Middle-incom

small-scaIIIb 5,100 19,390 Upper m

households ao

IVa 9,750 19,390 Large hotels, hbanks, a

IVb 11,500 27,665 Harb

Source: Pom Jaya and Thames Pam Jaya, personal commu

ta neighborhoods in 2005, and is supported byreferences to other published survey data. Thesurvey entailed a multi-stage sample, with thefirst stage using a clustered random sample to se-lect areal units with self-identified water accessproblems and a high proportion of poor house-holds, based on initial site visits, interviews withlocal officials (kapubaten), and consultationwith municipal and national government offi-cials. This resulted in the selection of eleven kel-urahan (sub-districts) in North (Kamal Muara,Penjaringan (Marlina & Gedong Kompa), Penj-aringan (Rawa Bebek), Pegangsaan Dua), West(Semanan), East (Kampung Melayu, Rawa Ter-ate, Jati, Kampung Tengah) and Central Jakarta(Kebon Melati, Gunung Sahari Selatan). Astratified random sample of 110 householdswas then conducted in these eleven kelurahan,with numbers of respondents proportionatelycorresponding to the overall distribution of pri-mary users of the four water sources in thesecommunities: stand-alone water ‘‘terminals’’ re-filled by tanker trucks by national governmentauthorities (50%), private deep wells (26%), sub-sidized house connections to the PAM Jaya net-work (17%), and public hydrants administeredby PAM Jaya (6%), based on data provided byKimpraswil (then-Ministry of Infrastructure/Public Works) through the water supply compo-nent of its Fuel Subsidy Reduction Compensa-tion Program. Qualitative and quantitativedata were collected and entered into a computer-ized database following transcription of the

new connections

ff groupn (2003–05)

Number of newconnections,East Jakarta(1998–2004)

Increase (%)

ons (e.g., reliqiouspublic hydrants

1,101 1

ls and very poorseholds

21,898 24

e householdse households andle businesses

51,847 58

iddle-incomend government

ffices

11,150 12

igh-rise buildings,nd factories

2,323 3

or/port 1,849 2total 90,167 100

nication.

Table 4. Water expenses, as a proportion of income(n = 110)

Income range (Rp.) Average (%) Maximum (%)

<750,000 14 96750,000–1,500,000 5 231,500,000–3,000,000 5 193,000,000–6,000,000 2 5>6,000,000 1 1

Source: Survey by authors of 110 households in six Ja-karta neighborhoods (2005).


qualitative data. While ensuring a uniform, com-prehensive, and randomized approach to sampleselection, 19 it should be noted that the deliber-ate choice of communities with self-identifiedproblems of water access was intended to docu-ment the governance failures affecting poorhouseholds in areas where some network supply(and hence at least the theoretical possibility ofconnecting to the network) did exist, rather thanbeing representative of all households across Ja-karta, many of which (both poor and wealthy)are in areas completely without water mains.

First, a brief summary of income and wateruse characteristics is presented. In the neigh-borhoods surveyed, a combination of some orall of the following water sources is available:in-house connections, public hydrants (individ-ual access or via water vendors), deep wells, and‘‘water terminals’’ (large stand-alone tanks re-filled by tanker trucks using networked watersupply). The majority of poor households usenon-networked water supply in the home (Ta-ble 3). Many use more than one type of watersources: according to our survey, 61% of house-holds surveyed used multiple sources (the threemost frequent combinations being network andvendor water, network and groundwater, andgroundwater and vendor water) 20 (Table 3).

The lower income households surveyed spent,on average, a greater proportion of their house-hold income on water supply (Table 4). This is inpart because of the greater proportional expenserepresented by water as a fraction of total house-hold income, but also because the choice of typesof water supply varies with income. Only 10% ofhouseholds in the lowest income bracket used

Table 3. Use of multiple water sources by Jakartahouseholds (n = 110)

Water source Number of houses %

Groundwater 39 37Groundwater with

bottled water/vendedwater/public hydrant

41 39

Network water 10 9Network water with

groundwater2 2

Public hydrant/vendedwater with rainwater

14 13

Total 106 100Total households using

at least two sources65 61

Source: Survey by authors of 110 households in six Ja-karta neighborhoods (2005).

networked water supply, whereas 30% of thosein the higher income bracket did so (Figure 2).Figure 3 indicates that those households relyingon groundwater, or a combination of groundwa-ter and networked water paid significantly less(in absolute terms, and as a proportion of in-come) for water supply than those householdsrelying on vendor water. Typically, these house-holds would own a well and associated storage(usually rooftop) infrastructure, implying a sig-nificant capital investment and relative perma-nency of residence. In contrast, householdsusing vendor water spent higher amounts (inabsolute rupiah terms), as well as higher relativeproportions of income; overall, 43% of house-holds spent more than 5% of their income onwater bills (often cited as an appropriate thresh-old by international aid organizations). 21 Yet,vendor water was consistently the highest costper unit volume in the six neighborhoods sur-veyed; the price per unit volume was found tobe from 10 times to 32 times more expensive thannetworked water. 22 In short, the behavior ofpoor households with respect to the choice ofwater services presents an apparent paradox:informal water services thrive in neighborhoodswhere formal services are available, with house-holds relying on water vendors even when theyhave the option of house connections with themunicipal water utility (Susantono, 2001).

Why would poor households make this choice?Analysis of household survey data indicates thatpoor households choose non-networked watersources for a variety of reasons, 23 including con-nection fees (total cost of water supply); transac-tion costs; housing and residence status; securityof water supply; and, in some instances, percep-tions of water quality. The first, and probablythe most important factor to consider is the totalcost of water supply (as distinct from the volumet-ric cost, or cost per unit volume of water). On thebasis of cost per unit volume alone, then, it seemscounter-intuitive that poor households would not

Figure 2. Income and type of water supply. Source: survey by authors of 110 households in six Jakarta neighborhoods

(2005).

Figure 3. Expenses on water supply. Source: survey by authors of 110 households in six Jakarta neighborhoods (2005).


connect to the water supply network where possi-ble. However, the disincentive for connection be-comes more obvious when we consider the totalcost of connecting to the water supply system(as opposed to price per unit volume of water sup-ply). Monthly bills include more than charges perunit volumes of water consumed. Fixed chargessuch as the meter fee and the annual charge (dis-tinct from the volumetric charges discussedabove) are often orders of magnitude higher thanthe volumetric price per cubic metre and are alsoadded on to the bill (Table 2). For a poor house-hold whose residents consume 50 L/person/day(the World Health Organization recommendedminimum), fixed charges will be anywhere from5 to 10 times as high as the volumetric consump-tion charge; the effective cost per unit volume will

thus be higher than that of vendor water for thepoorest consumers. However, it is also likely thathouseholds do not have access to sufficient infor-mation to assess the cost differences between dif-ferent water supply options: most householdswere not aware of the fact that volumetric costsof water supply through an individual house con-nection were less expensive than alternativesources.

Household transaction costs are also signifi-cant. A networked water supply implies addi-tional infrastructure costs to be borne by theconsumer, in the form of a water tank or hold-ing device, made necessary because of the inter-mittent nature of water supply through thepiped network (with cutoffs of several hoursoccurring daily in some areas). Connection fees


alone may be prohibitive (ranging from 200,000to 350,000 Rupiah in the households surveyed,with reported figures from other surveys some-times much higher than this), relative to averageincomes of poor households (which averaged1.4 million Rupiah/month in the householdssurveyed), and must usually be provided as alump sum, which may pose significant barriersto households with small, irregular incomes.Connection fees also vary depending on dis-tance from the network; poor households aremore likely to live in areas of lower networkdensity, and thus are more likely to pay higherfees for connecting. Lower density of networkconnections in poor neighborhoods (as docu-mented in Section 2) thus translated into anadditional barrier for poor customers wishingto connect to the network. Finally, long waitingtimes at water utility offices to pay bills and dealwith meter mis-readings raise transaction costscompared to the ease of complaint handlingand convenience of home visits by vendors tocollect bill payments. Moreover, payment flexi-bility permitted by vendors (some of whom evenallow customers to buy water on credit) pro-vides an important incentive for poor house-holds, which may have limited budgetingability (Shofiani, 2003; Susantono, 2001).Although difficult to estimate, bribes demandedby contractors and water utility staff (or theapprehension that such bribes may bedemanded) potentially represent anotherimportant transaction cost. 24 These transac-tions costs are likely harder to bear for thosehouseholds with variable income: asFigure 4 indicates, households with fluctuatingincomes are more likely to rely solely on vendorwater. For all of these reasons, overall costs ofvendor water to poor households may be lowerthan networked water supply, even though thelatter has a lower price per unit volume.

A third factor is tenure, or housing and resi-dency status. A significant proportion of the ci-ty’s population (70%, in our sample) lives inrental or temporary (often self-built) accommo-dation without secure tenure. Deep wells areexpensive and have higher maintenance costs,effectively prohibiting development by thosewithout permanent tenure. Moreover, land-lords are often unwilling to connect rentalproperties to the water supply network becauseof concerns about infrastructure cost and main-tenance; similarly, tenants are unwilling to con-nect, because their investment would constitutean upgrade to the landlord’s property. In oursurvey, households with insecure tenure were

significantly less likely to have a householdwater supply connection: 32% of householdssurveyed owned their home and possessed landcertificates, and 100% of these households re-lied on networked water supply as their pri-mary source, whereas the remaining 68%(tenants, or owners without land certificates)used a combination of vendor water and shal-low groundwater. In addition, a large numberof city residents are without legal residency per-mits and/or official (registered) land permits,and are subject to a complex, variable, andunstable land tenure regime which creates sig-nificant disincentives for poor households toobtain official registration and associated per-mits (Leaf, 1994; Server, 1996; Struyk,Hoffman, & Katsura, 1990). These are addi-tional reasons for which water vendors—whichprovide flexible, easily accessible, low transac-tion cost, ‘‘no questions asked’’ water sup-ply—may be preferable.

Water availability may also factor in the deci-sion not to connect to the water supply network.Low pressure in poorer areas of the city (Map 1)implies that water flow is intermittent, oftenresulting in greater infiltration of pollutants intothe network, thereby reducing water quality.Low pressure in the piped network also meansthat households prefer to have a backupsource—usually a shallow well. During the dryseason, water vendors represent a more securesource of water for households than the network.

Finally, perceptions of water quality play animportant role in households’ decision-making.Water vendors were perceived by survey respon-dents (particularly those with more education)to supply water of higher quality, whereas otherrespondents perceived groundwater to be ofhigher quality. These perceptions may havesome factual basis: the most comprehensivecomparative survey of water quality of differentsources in poor neighborhoods in Jakarta todate found that samples of drinking water fromthe network were more contaminated with fecalcoliform than groundwater (Surjadi et al., 1994).The fact that vendors check water quality andmay strain the water or let it settle before deliver-ing explains why perceptions of vendor waterquality may be higher, despite the fact that ven-dor water often originates in hydrants connectedto the networked water supply system. Theseperceptions of relative water quality of differentsources may not, however, be borne out in allcases; groundwater quality tends to be lower inpoorer areas, due to proximity to industrial sites,fewer controls over sewage, density, and higher

Figure 4. Type of income and type of water supply. Source: survey by authors of 110 households in six Jakarta

neighborhoods (2005).


salinity levels (particularly in North Jakarta).Dislike chlorine (as borne out by the widespreadpractice of leaving buckets of water to sit over-night to dissipate the chemical smell) is anothercontributing factor (Kreimer et al., 1995;McGranahan et al., 2001). Echoing the concernsof survey respondents, generalized public mis-trust of the quality of network water in Jakartais evident from the rapid growth of bottled waterconsumption by poor households in Jakartaover the past five years, particularly via ‘‘no-name’’ air isi ulang (refilled bottled water)supplied by private (and unregulated) watertreatment micro-plants throughout the city(Forkami, 2006; Weimer, 2006). 25

These governance failures were not redressedby the introduction of private sector manage-ment of Jakarta’s water supply network in1998. Land tenure status and regulations fallbeyond the purview of the water supply utility,whether under public or private management.The failure of the World Bank’s ‘‘Output BasedAid’’ project in Jakarta to meet targets is anexample of the barriers posed by land tenure,and the inability of the water supply utility toresolve these issues. Under the OBA project,20,000 households were targeted to receivenew, subsidized water supply connections; how-ever, most households qualifying as sufficientlypoor to receive a subsidized connection do nothave legal land tenure status, precluding theirparticipation in the programme. Despite lobby-ing of government by the private sector opera-tors to change this tenure policy, the originaltarget has been scaled back to less than 4,000households (Forkami, 2006).

Issues related more directly to water supply—higher transaction costs for poor households,perceptions of water quality and availability,

and security of water supply—could perhapshave been more feasibly addressed by the privateconcessionaires, and indeed some changes havebeen made. One of the two private concession-aires, for example, began allowing householdsto pay connection fees in instalments over a 12-month period rather than demanding full pay-ment up-front. But other significant issues,including low water pressure in poor neighbor-hoods, the overall burden of connection feesand transaction costs, the disjuncture betweencurrent billing practices and the ability-to-payof poor households with fluctuating incomes,and water quality (particularly the admitted lackof potability, exacerbated by the fact that the ori-ginal PSP contract target of achieving potablewater standards was dropped in subsequentnegotiations) were not addressed by the privateconcessionaires. Indeed, the failure to extendwater supply connections to poor neighbor-hoods in order to meet contractual targets wasa factor cited by Thames Water in its decisionto withdraw from the contract in late 2006.

6. CONCLUSIONS

This paper has demonstrated that Jakarta’swater supply system has been highly frag-mented since its inception: access to a house-hold network connection has been stronglydifferentiated economically (i.e., poverty is cor-related with lack of access to a household con-nection, with the use of alternative watersources, with low levels of water consumption,and with spending higher proportions of house-hold income on water supplies) and spatially(i.e., those lacking access are concentrated inspecific districts of the city, and within lower in-


come areas in neighborhoods across the city).Section 2 argued that this differentiation of ac-cess has deep historical roots. The current lackof access throughout large areas of Jakarta isdue, in part, to the legacy of segregated colonialwater supply systems, and deliberate under-investment in the post-colonial period, as pol-icy-makers sought to discourage rural–urbanmigration, and gave low priority to extendingwater supply access to the urban poor, focusinginstead on economic development of key sec-tors, or on an urban redevelopment agenda fo-cused on ‘‘monumental’’ infrastructure.

This analysis reminds us that we should bewary of viewing cities such as Jakarta througha Northern lens. In many cities in the globalSouth, fragmentation of utility services suchas water is not due to the recent trends of‘‘splintering urbanism’’ characteristic of citiesin the North (Graham & Marvin, 2001), butrather due to a model of urbanization withroots in the colonial era which produces a per-sistent pattern of differentiation of services andurban spaces, under public utility management.Post-independence ‘‘state failure’’ (or, to useDavis’ term, the ‘‘betrayal of the state’’ (Davis,2006)) was in part shaped by the technologicalconstraints embodied by the colonial era waternetwork, and then exacerbated by the captureof public sector organizations by elite interests.

Sections 3 and 4 attempted to explain whythis situation arose and persists, and presentedevidence on governance failures affecting urbangovernment and the water supply utility, as wellas households. Framing the analysis of theseinstitutions and incentives as governance fail-ures usefully reminds us that an understandingof choices by both utilities and households isnecessary in analyzing the failure to achieveuniversal urban water supply provision. In thecase of Jakarta, the culture of urban gover-nance, the conventional water supply utilitybusiness model adopted in Indonesia, broaderurban planning constraints, cost recoveryrequirements, and economic incentives linkedto tariff structures were identified as some ofthe contributing factors to governance failure.The factors preventing or precluding the capa-bility of poor households to connect to the net-work include limited ability-to-pay of totaltransaction costs (rather than volumetric costs);insecure tenure; the inability of the water sup-ply utility to deal with poor households’ needfor flexible payment options; and perceptionsof the relative quality, availability, and reliabil-ity of different water sources. These governance

failures arose under public management, buthave not been significantly remedied since theintroduction of private sector management ofthe city’s water supply system in 1998.

These findings lend support to the recentstudies that suggest that institutional and gover-nance issues should receive greater attention inthe pro-poor water supply debate, and that thedebate over the relative merits of public and pri-vate sector provision has diverted attentionfrom the pressing issue of governance reformnecessary to meet water-related developmentgoals. (e.g., ADB, 2003a; Bayliss & Fine, 2007;UNCHS, 2003, 2006 UNDP, 2006; UNWWAP,2006; WaterAid, 2003; Whittington, 1992). Spe-cific institutional and governance issues requireattention, namely: the institutions of urban gov-ernance (which, in Jakarta’s case, has systemat-ically prioritized monumental infrastructureand elite residential services at the expense ofuniversal public services); the inequitable spa-tialization of network access; and the multipledisincentives for the poor to choose networkconnections, and for network managers—bothpublic and private—to connect the poor. Inturn, these recommendations lend weight tothe growing body of ‘‘good governance’’ proto-cols (e.g., Gomez-Lobo, 2001; Jansky & Uitto,2005; UNCHS, 2006) that focus on institutionsand incentives as central to improving utilityperformance, particularly with respect to urbanwater supply for the poor.

This, in turn, suggests several policy implica-tions with respect to pro-poor water supplydelivery. First, governance failure (rather thanthe putative failures or merits of public or pri-vate providers) should be the focus of policymakers. For example, utilities—both publicand private—should have clear benchmarksfor equity associated with non-compliance pen-alties; in some cases, these benchmarks couldhave a spatial component (e.g., numbers ofconnections within specific neighborhoods orsupply zones). And all water supply providersshould be subject to robust regulatory frame-works with clear standards for good gover-nance (such as accountability, transparency,participation, inclusiveness, and the rule oflaw), which should be implemented for watersupply systems—whether public or private.

Second, we need to realistically reconsider the‘‘modern infrastructural ideal’’ (which envisionedthe universal extension of uniformly regulatednetworks) (Graham & Marvin, 2001). Currently,networked water supply utilities tend to be sub-ject to some form of regulation, whereas the


non-networked water supply alternatives uponwhich a large proportion of urban poor house-holds depend are often entirely unregulated. Incontrast, we suggest that the scope of regulationand public oversight should not be constrainedto networked water supply provision, but ratherinclude the diverse array of water provision sys-tems actually operating in urban environments(with, e.g., co-defined but distinct service stan-dards and water quality requirements for differenttypes of providers). Moreover, development poli-cies which support alternatives should be basedon sound governance principles such as equitableaccess, transparency, and public health protec-tion. This is particularly relevant given the largenumber of alternatives that are being proposedin the current debate, including business models(e.g., public–public partnerships), financing(e.g., municipal bond finance), governance mod-els (such as communal water rights), and technol-ogies (such as condominal water supply) are allcurrently being explored as viable options for sus-tainable urban water supply (e.g., Katko, 2000;Kay, 1996; Narain, 2006; PSIRU, 2006; Trawick,2003; TNI, 2005, 2006; Zaidi, 2001).

These recommendations will, we suggest, beuseful in supporting the fulfillment of the waterand sanitation Millennium Development Goal,which will contribute to other MillenniumDevelopment Goals (such as those related tochild health) because of the positive externali-ties associated with water supply provision(Fay, Leipziger, Wodon, & Yepes, 2005). Insupporting or developing these alternative op-tions, it is, of course, critical to avoid the crea-tion of ‘‘two-tier’’ systems, in which onlywealthy households have access to adequatewater supply. But it is also important toacknowledge the limitations and unsuitabilityof conventional networked water supply sys-tems for many urban environments and resi-dents. Rather than excluding these alternativesystems from consideration, they should ac-tively be assessed and regulated—properlyexpanding the role of state regulation and,where appropriate, operational managementand finance, to include all water supply activi-ties within urban environments.

NOTES

1. The World Health Organization and UNICEF JointMonitoring Programme for Water Supply and Sanitationprovided figures for the proportion of the population with‘‘improved’’ provision, including water from publicstandpipes, boreholds, and protected dug wells, providedthat at least 20 L per person per day is available from asource within 1 km of the person’s home. This definitiondoes not account for whether the water is ‘‘adequate’’ or‘‘safe’’ to drink (WHO/UNICEF, 2005).

2. Water supply needs in rural areas are very distinctfrom those of urban areas; hence, the analysis isconstrained to urban populations.

3. Initial research took place in Jakarta in 2001,followed by archival work in the colonial archives inthe Netherlands in 2003. Subsequent research took placeover an 18-month period from 2004 to 2005. Data wereobtained from: (1) a household survey of 110 poorhouseholds in six Jakarta ‘‘kampungs’’ (neighborhoods)in 2005; (2) public and internal reports from the twoprivate concessionaires, the Jakarta Water Supply Reg-ulatory Body, and the Jakarta municipal government;(3) over 60 interviews with water supply managers,government officials, international financial institutions,aid agencies, and NGO representatives in 2001, 2004 and2005; (4) a detailed archival survey at KITLV (Kon-

inklijk Instituut voor Taal-, Land- en Volkenkunde/Royal Netherlands Institute of Southeast Asian andCaribbean Studies) and KIT (Koninklijk Instituut voorde Tropen/Royal Tropical Institute) in Leiden andAmsterdam, The Netherlands, in 2003.

4. Institutions are here defined in as the laws, regula-tions, norms, and customs which structure socialbehavior, together with associated incentives (Ostrom,1990; Saleth & Dinar, 1990). Ownership is defined asproperty rights, together with the organizations (e.g.,public sector, private sector, or community-based) thatpossess those rights. Institutions are thus analyticallydistinct from organizations, although in practice the twoare inter-related and often co-evolve (North, 1995).

5. The term ‘‘institution’’ is used in the sociologicalsense: laws, rules, norms, customs, and incentivesgoverning behavior and decision-making.

6. Collignon and Vezina (2000) identified six types ofprivate sector operators: concessionaires; pump opera-tors; ‘‘carters’’ and ‘‘carriers’’ (ambulatory vendors);standpipe managers; repair companies; and maintenancecontractors. Only the first four would typically under-take direct sales to consumers.


7. The first figure is from the annual SUSENAS socio-economic survey conducted by the Indonesia Bureau ofStatistics (BPS, 2005). The second figure was calculatedusing data from the regulatory authority overseeing themanagement of Jakarta’s water supply system (JakartaWater Supply Regulatory Body, 2004). This was cross-referenced with an ADB report (2003b) that reported afigure of 51.2%. In Jakarta, coverage ratios are alwaysimprecise estimates; their calculation is dependent upona number of variables that are only imprecisely mea-sured, such as urban population and average size ofhousehold. Reported figures vary significantly, and donot indicate the number of households which have aconnection, but indicate which rely primarily on othersources (e.g., groundwater) due to quality or serviceconcerns (e.g., low pressure). Large numbers of seasonalmigrants and ‘‘illegal’’ residents without land tenuremean that population figures are systematically under-estimated and that, as a consequence, coverage figuresare systematically inflated.

8. Kampung translates literally as ‘‘village,’’ but wasused to refer to non-European, and in particularIndonesian neighborhoods of Batavia, thereby demar-cating the European sections of the city as the sole‘‘urban’’ zones.

9. Specifically, in 1929, 6,926 kampung households(estimated population 400,000) were supplied with 24 L/s, while 10,392 European households, a population of37,067, were supplied with 84 L/s (Eggink, 1930).

10. Over 60 interviews were conducted with govern-ment (national, regional and local government repre-sentatives), aid agencies (NGOs, bilateral aid agencies,and multilateral financial institutions), and water sup-pliers (public and private).

11. DKI Jakarta alone grew from an estimated 1.8million people in 1950 to an estimated 6.5 million in1980, with equally rapid population growth in thesurrounding metropolitan areas, particularly at theexpanding rural–urban fringe beyond the boundariesof DKI Jakarta (Chifos, 2000; Firman, 1997, 1998, 2000;Lo & Yeung, 1996).

12. Notably through the Kampung Improvement Pro-gram (KIP) (1966–77), but this produced only marginalimprovements in the access of lower income residents topiped water (Abeyasekere, 1987; KIP, 1976; Taylor,1983; World Bank, 1974).

13. Kreimer, Gilbert, Volonte, & Brown, 1995; Taylor,1983; World Bank, 1974 recorded how the ‘‘basic needs’’development programs in the 1970s and 1980s thatintended to provide ‘‘water for the poor’’ through public

hydrants were often frustrated due to cost recoveryrequirements. Previously, however, this had not been thecase. Up until the 1960s, consumption was not metered,and rates were low; water was essentially distributed freeto consumers (largely in wealthier neighborhoods).Ironically, low cost recovery at this point preventedthe accumulation of sufficient capital to enable expan-sion into poorer neighborhoods; cost recovery was attimes so low that PAM Jaya was often unable to pay itsemployees (PAM JAYA, 1992a).

14. For water supply utilities (PDAMS) across Indo-nesia, the standard level of this dividend was 55% of netprofits, which is paid to the treasury of the localmunicipality (World Bank, 1997b). However, this ismandatory only when 75% coverage of the population isreached (for PDAMs in urban areas); moreover, inpractice, all or part of this dividend is frequentlyreturned to the utility.

15. The use of ‘‘user-fees’’ to finance network extensionmeant that house connections were unaffordable for themajority of the population; the cost of a householdconnection (not to mention ‘‘additional fees,’’ meterrental, deposit, and actual monthly tariffs) in 1975 wasRp. 100,000 ($200 US), whereas the average income inJakarta at that time was only Rp. 15,000/mth (approx-imately $36 US), with the 80% of the city’s residentsliving in kampungs earning much less than that amount(KIP, 1976; PERPAMSI, 1975a).

16. The management of Jakarta’s water supply systemwas outsourced to two private sector consortia, ThamesWater (east Jakarta) and Ondeo (Lyonnaise des Eaux)(west Jakarta), in 1998. Reputedly due to continuedlosses, Thames Water sold its stake in TPJ (its Jakartawater services subsidiary) to an Indonesian-led consor-tium in 2006 (Jakarta Post, 2006).

17. Until the early 1980s, household tariffs were 25Rp./m3 for the first 15 m3/mth; public hydrants andwater trucks paid Rp. 60/m3, more than double the tariffof households, and more than even small businesses(who paid 50 Rp./m3). (PERPAMSI, 1975a, 1975b,1975c).

18. The cumulative deficit by the end of 2001 was Rp.469 billion (approximately $46 million USD) and hadreached Rp. 990 billion (approximately $97 millionUSD) by September 2003—excluding late paymentinterest and retroactive tariff increases (Jakarta WaterSupply Regulatory Body, 2005).

19. The survey controlled for gender bias by selectingonly female respondents (on the basis of the significantlyhigher responsibility for women for all water-related


activities within the home). The average householdmonthly income of 1.4 million Rupiah (with an averageof 5.7 household residents) translates into an averageindividual income of 245,614 Rupiah, slightly above the2005 official poverty line, but well below the level ofincome earned by minimum wage earners.

20. These findings are similar to the results of surveysconducted by Surjadi et al. (1994, 2002, 2003) andMcGranahan et al. (2001), the two most recent academicstudies available.

21. A study of 1000 households in Jakarta whichexamined the different prices paid by different wealthgroups found that, overall, the poor pay on averagetwice as much per metre cubed as the wealthy(McGranahan et al., 2001), and that water expenditurerepresents, on average, 10% of income in poor house-holds.

22. ADB (2003b); McGranahan et al. (2001); and asurvey conducted by the author in the neighborhood ofSunter Agung in January 2001. ADB 2003b gave amaximum figure of US $4.17/m3.

23. For a history of the preferences of poor householdsfor alternative water supplies, see: Chifos, 2000; Kreimeret al., 1995; Surjadi, Padhmasutra, Wahyuningsih,McGranahan, & Kjellen, 1994; Taylor, 1983; YayasanDian Desa, 1990.

24. See Cowherd, 2002 for a discussion of the cultureof ‘‘informal’’ profits from public services, and (YayasanDian Desa, 1990) for how this was evident in PAMJaya’s operations in poorer communities.

25. ‘‘Air isi ulang’’ sells for Rp. 3,000/19 L; ‘‘brandname’’ bottled water bought by middle/upper classhouseholds typically retails for Rp. 10,000/19 L(Weimer, 2006).

REFERENCES

Abeyasekere, S. (1987). Death and Disease in 19thCentury Batavia. Death and disease in SoutheastAsia: Explorations in social, medical, and demo-graphic history. N. Owen. Singapore, Oxford, NewYork: Oxford University Press.

Abeyasekere, S. (1989). Jakarta: A history. Singapore:Oxford University Press.

ADB (2003a). Beyond boundaries: Extending services tothe urban poor. Manila: Asian Development Bank.

ADB (2003b). Water in Asian cities: Summary of findingsof the study and a regional consultation workshop.Manila: Asian Development Bank.

Agtini, M., Rooswanti, S., Lesmana, M., Punjabi, N.,Simanjuntak, C., Wangsasaputra, F., et al. (2005).The burden of diarrhoea, shigellosis, and cholera inNorth Jakarta, Indonesia: Findings from 24 monthssurveillance. Infectious Disease, 5(89), 1–2.

Anwar, A. (2001). Regulating services for the poor,Jakarta, Indonesia. Presented at the 3rd WorldWater Forum, Osaka, 19th March. Jakarta WaterSupply Regulatory Board, Jakarta.

Argo, T. (1999). Thirsty downstream: The provision ofclean water in Jakarta, Indonesia. Unpublisheddoctoral thesis, School of Regional and CommunityPlanning, University of British Columbia.

Argo, T., & Firman, T. (2001). To privatize or not toprivatize? Reform of urban water supply services inJabotabek, Indonesia. Built Environment, 27(2),146–155.

Azdan, M. (2001). Water policy reform in Jakarta,Indonesia: A CGE analysis. Unpublished doctoralthesis. Department of Agricultural, Environmentaland Development Economics, Ohio State University.

Bakker, K. (2003). From archipelago to network:Urbanisation and privatisation in cities in the South.The Geographical Journal, 169(4), 328–341.

Bakker, K. (2007). Trickle down? Private sector partic-ipation and the pro-poor water supply debate inJakarta, Indonesia. Geoforum, 38(5), 855–868.

Bayliss, K. (2002). Privatisation and poverty: Thedistributional impact of utility privatisation. Annalsof Public and Co-operative Economics, 73(4),603–625.

Bayliss, K. (2003). Utility privatisation in Sub-SaharanAfrica: A case study of water. Journal of ModernAfrican Studies, 41(4), 507–531.

Bayliss, K., & Fine, B. (2007). Privatization and alter-native public sector reform in Sub-Saharan Africa:Delivering on electricity and water. London: Pal-grave MacMillan.

Berry, B. (1982). Clean water for all: Equity-based urbanwater supply alternatives for Indonesia’s cities.Urban Geography, 3(4), 281–299.

Bond, P. (2002). Unsustainable South Africa: Environ-ment, development and social protest. London:Merlin.

Bond, P. (2004). Water commodification and decom-modification narratives: Pricing and policy debatesfrom Johannesburg to Kyoto to Cancun and back.Capitalism Nature Socialism, 15(1), 7–25.

BPD (2003). Jakarta, Indonesia: Water supply improve-ments to Marunda District. London: BuildingPartnerships for Development in Water and Sanita-tion. <http://www.bpd-waterandsanitation.org/eng-lish/projects/jakarta.asp>. Accessed 3.01.04.

BPS (2005). Statistik Kesejahteraan Rakyat/Social wel-fare statistics. Jakarta: Indonesian Bureau of Statis-tics.

Braadbaart, O. (2002). Private versus public provision ofwater services: Does ownership matter for utilityefficiency?. AQUA – Journal of Water Supply:Research and Technology, 51(7), 375–388.

http://www.bpd-waterandsanitation.org/english/projects/jakarta.asp

http://www.bpd-waterandsanitation.org/english/projects/jakarta.asp


Braadbaart, O., & Braadbaart, F. (1997). Policing theurban pumping race: Industrial groundwater over-exploitation in Jakarta. World Development, 25(2),199–210.

Brennan, E. M., & Richardson, H. W. (1989). Asianmegacity characteristics, problems and policies.International Regional Science Review, 12(2),117–129.

Brown, A. (2001). Confusing means and ends: Frame-work of restructuring, not privatization, mattersmost. International Journal of Regulation and Gover-nance, 1(2), 115–128.

Budds, J., & McGranahan, G. (2003). Are debates onprivatization missing the point? Experiences fromAfrica, Asia, and Latin America. Environment &Urbanization, 15(2), 87–113.

Cairncross, S., Hardoy, J., & Satterthwaite, D. (Eds.). The poor die young. London: Earthscan.

Cestti, R., Batia, R., & Van der Berg, C. (1994). Waterdemand management and pollution control in theJabotabek region, Indonesia. Unpublished WorldBank paper, October.

Chifos, C. (2000). Meeting unmet expectations: Initia-tives in urban environmental service provision inJakarta. In C. Chifos, & R. Yabes (Eds.), SoutheastAsian urban environments. Program for SEA studies(pp. 29–72). Arizona State University: Temple Uni-versity Press.

Chifos, C., & Hendropranoto, S. (2000). Thirty years ofurban infrastructure development in Indonesia. In C.Chifos, & R. Yabes (Eds.), Southeast Asian urbanenvironments. Program for SEA studies(pp. 153–181). Arizona State University: TempleUniversity Press.

Collignon, B., & Vezina, M. (2000). Independent waterand sanitation providers in African cities: Full reportof a ten-country study. Washington: World Bank.

Cowherd, R. (2002). Cultural construction of Jakarta:Design, planning and development in Jabotabek,1980–1997. Unpublished PhD dissertation, Depart-ment of Architecture, MIT.

Crane, R. (1994). Water markets, market reform and theurban poor: Results from Jakarta, Indonesia. WorldDevelopment, 22(1), 71–83.

Crane, R., Daniere, A., & Harwood, S. (1997). Thecontribution of environmental amenities to low-income housing: A comparative study of Bangkokand Jakarta. Urban Studies, 34(9), 1495–1512.

Cross, P., & Morel, A. (2005). Pro-poor strategies forurban water supply and sanitation services deliveryin Africa. Water Science & Technology, 51(8), 51–57.

Davis, M. (2006). Planet of slums. London: Verso.Eggink, E. J. (1930). ‘Na 25 Jaar’. Beknopt Gedenksch-

rift ter Gelegenheid van het 25 Jarig Bestaan derGemeente Batavia (‘After 25 Years’. A concisecommemorate at the occasion of 25 years existenceof the Municipality of Batavia) (Gemeente Batavia,Batavia).

Estache, A., & Rossi, M. (2002). How different is theefficiency of public and private water companies inAsia? The World Bank Economic Review, 16(1),139–148.

Fay, M., Leipziger, D., Wodon, Q., & Yepes, T. (2005).Achieving child-health-related millenium develop-ment goals: The role of infrastructure. World Devel-opment, 33(8), 1267–1284.

Firman, T. (1997). Land conversion and urban devel-opment in the northern region of West Java, Indo-nesia. Urban Studies, 34, 1027–1046.

Firman, T. (1998). The restructuring of Jakarta metro-politan area: A ‘‘global city’’ in Asia. Cities, 15(4),229–243.

Firman, T. (2000). Rural to urban land conversion inIndonesia during boom and bust periods. Land UsePolicy, 17, 13–20.

Firman, T., & Dharmapatni, I. (1994). The challenges tosustainable development in Jakarta metropolitanregion. Habitat International, 18(3), 79–94.

Fischer, L. (1959). The story of Indonesia. New York:Harper Brothers.

Ford, L. (1993). A model of Indonesian city structure.Geographical Review, 83(4), 374–396.

Forkami/RTI (2002). Update: Report on water supplyto the urban poor (2002). Forkami and RTI Inter-national, Jakarta.

Forkami/RTI (2003). Update: Report on water supplyto the urban poor (2003). Forkami and RTI Inter-national, Jakarta.

Forkami (2006). Social assessment for the jakarta output-based aid project. Jakarta: Forum Komunikasi Pen-gelolaan Kualitas Air Minum Indonesia.

Franceys, R., & Weitz, A. (2003). Public-private com-munity partnerships in infrastructure for the poor.Journal of International Development, 15(8),1083–1098.

Gilbert, A., & James, D. (1994). Water pollution inJakarta Bay. In D. James (Ed.), The application ofeconomic techniques in environmental impact assess-ment (pp. 111–142). Dordrecht: Kluwer.

Gleick, P., Wolff, G., Chalecki, E., & Reyes, R. (2002).The new economy of water: The risks and benefits ofglobalization and privatization of fresh water. Oak-land: Pacific Institute.

Gomez-Lobo, A. (2001). Incentive based subsidies:Designing output-based subsidies for water con-sumption. Public policy for the private sector 22.Washington: World Bank.

Graham, S., & Marvin, S. (2001). Splintering urbanism.London: Routledge.

Gulyani, S., Talukdar, D., & Kariuki, R. M. (2005).Universal (Non)service? Water markets, householddemand and the poor in urban Kenya. UrbanStudies, 42(8), 1247–1274.

Gutierrez, E., Calaguas, B., Green, J., & Roaf, V. (2002).New rules, new roles: Does PSP benefit the poor?London: WaterAid and Tearfund.

Harsono, A. (2005). When water and political powerintersect: A journalist probes the story of waterprivatization in Jakarta, Indonesia. Nieman Reports,59(1), 45–47.

Hodge, G. (2000). Privatization: An international reviewof performance. Boulder: Westview.

Holmes, P. (2000). Effective organizations for watermanagement. International Journal of WaterResources Development, 16(1), 57–71.


Hukka, J. J., & Katko, I. (2003). Refuting the paradigmof water services privatization. Natural ResourceForum, 27, 142–155.

Hunt, L. C., & Lynk, E. L. (1995). Privatisation andefficiency in the UK water industry: An empiricalanalysis. Oxford Bulletin of Economics and Statistics,57, 371–388.

Ibrahim, S. (2006). From individual to collective capa-bilities: The capability approach as a conceptualframework for self-help. Journal of Human Develop-ment, 7(3), 397–416.

Jakarta Post (2006). PT Aquatico to take over city waterconcession. November 27th.

Jakarta Water Supply Regulatory Body (2004). Regula-tory approach to the Jakarta water supply concessioncontracts. Presentation: A. Lanti (Member, Regula-tory Body), Stockholm water symposium, August16–20.

Jakarta Water Supply Regulatory Body (2005). Urbanwater in Jakarta. Presentation: Urban water andcitizenship workshop, University of Indonesia, 11May.

Jansky, L., & Uitto, J. (2005). Enhancing participationand governance in water resources management:Conventional approaches and information technology.Water resources management and policy series. NewYork: United Nations University Press.

JICA (1997). Study: Jakarta water supply developmentproject. Final report. Vol. 2, Main report. May 1997.Nihon Suido Consultants Co., Ltd and Nippon KoeiCo., Ltd.

Johnstone, N., & Wood, L. (2001). Private Firms andPublic Water: realising social and environmentalobjectives in developing countries. Cheltenham:Edward Elgar.

Katko, T. (2000). Water! Evolution of water supply andsanitation in Finland from the mid-1800s to 2000.Tampere: Finnish Water and Waste Water WorksAssociation.

Kay, J. (1996). Regulating private utilities: The customercorporation. Journal of Co-operative Studies, 29(2),28–46.

KIP (1976). Jakarta’s Kampung improvement pro-gramme. Directorate of Physical Development,Jakarta Capital City Government and KIP Imple-mentation Unit, Jakarta.

Kirkpatrick, C., Parker, D., & Zhang, Y. F. (2006). Anempirical analysis of state and private-sector provi-sion of water services in Africa. World BankEconomic Review, 20(1), 143–163.

Kreimer, A., Gilbert, R., Volonte, C., & Brown, G.(1995). Indonesia impact evaluation report. Enhanc-ing the quality of life in Urban Indonesia: The legacyof the Kampung improvement program, June 29.Operations Evaluation Department, Rep.14747-IND, Washington, DC: World Bank.

Kusno, A. (1997). Modern beacon and the traditionalpolity: Jakarta in the time of Sukarno. Journal ofSoutheast Asian Architecture, 1, 25–45.

Kusno, A. (2000). Behind the postcolonial: Architecture,urban space and political cultures in Indonesia. Lon-don: Routledge.

Leaf, M. (1994). Legal authority in an extralegal setting:The case of land rights in Jakarta, 1960–65. IndonesiaJournal of Planning Education and Research, 14(1),12–18.

Leclerc, J. (1993). Mirrors and the lighthouse: A searchfor meaning in the monuments and great works ofSukarno’s Jakarta. In P. Nas (Ed.), Urban symbolism(pp. 38–58). Leiden: E.J. Brill.

Leitmann, J. (1995). Urban environmental profile: Aglobal synthesis of seven urban environmental pro-files. Cities, 12(1), 23–39.

Lobina, E., & Hall, D. (2000). Public sector alternativesto water supply and sewerage privatization: Casestudies. International Journal of Water ResourcesDevelopment, 16(1), 35–55.

Lovei, L., & Whittington, D. (1993). Rent-extractingbehaviour by multiple agents in the provision ofmunicipal water supply: A study of Jakarta, Indo-nesia. Water Resources Research, 29(7), 1965–1975.

Lo, F. C., & Yeung, Y. M. (1996). Global restructuringand emerging urban corridors in Pacific Asia. In Y.M. Yeung, & F. C. Lo (Eds.), Emerging world citiesin Pacific Asia. Tokyo: United Nations UniversityPress.

MacDonald, G. (1995). Indonesia’s Medan Merdeka:National identity and the built environment. Anti-pode, 27(3), 270–293.

Maronier, V. F. C. (1929). De Drinkwatervoorzieningvan Batavia (The drinking water network of Bata-via). De Waterstaats-Ingenieur, 8, 223–239.

Martijn, E. J. (2005). Hydraulic histories of Jakarta,Indonesia (1949–97): Development of piped drinkingwater provision. Vancouver: Unpublished researchreport, Department of Geography, University ofBritish Columbia.

Martin, N. (2004). Corporatization as a means ofimproving water quality: The experience in Victoria,Australia. Journal of Toxicology and EnvironmentalHealth, Part A, 67, 1888–1899.

McDonald, D., & Ruiters, G. (2005). The age ofcommodity: Privatizing water in Southern Africa.London: Earthscan.

McGranahan, G., Jacobi, P., Songsore, J., Surjadi, C., &Kjellen, M. (2001). The citizens at risk: From urbansanitation to sustainable cities. London: Earthscan.

McIntosh, A. (2003). Asian water supplies: Reaching theurban poor. Manila: Asian Development Bank andInternational Water Association.

Narain, S. (2006). Community-led alternatives to watermanagement: India case study. Background paper:Human development report 2006. United NationsDevelopment Programme, New York.

Newbery, D. (2000). Privatization, restructuring, andregulation of network utilities. Cambridge: MITPress.

Nickson, A., & Franceys, R. (2003). Tapping the market:The challenge of institutional reform in the urbanwater sector. London: Palgrave MacMillan.

North, D. (1995). The new institutional economics andthird world development. In J. Hunter, J. Harris, &C. Lewis (Eds.), The new institutional economics andthird world development (pp. 17–26). London: Routl-edge.


Nunan, F., & Satterthwaite, D. (2001). The influence ofgovernance on the provision of urban environmentalinfrastructure and services for low-income groups.International Planning Studies, 6(4), 409–426.

Ohlsson, L. (2000). Water conflicts and social resourcescarcity. Phys. Chem. Earth (B), 25(3), 213–220.

Ostrom, E. (1990). Governing the commons: The evolu-tion of institutions for collective action. CambridgeUniversity Press.

Ostrom, E., Schroeder, L., & Wynne, S. (1993). Institu-tional incentives and sustainable development: Infra-structure policies in perspective boulder. WestviewPress.

PAM JAYA (1992a). PAM DKI 1922–1992. In PAMJAYA (1992) (pp. 13–29), Jakarta, Indonesia.

PAM JAYA (1992b). Usaha Peningkatan PelayananMasyarakat (Efforts to Improve Services for thePublic), Ir. H. Poedijono (Direktur Usaha). In PAMJAYA (1992) (pp. 34–39), Jakarta, Indonesia.

Pandjaitan, G. (2004). Kondisi Pelayanan Air Minum diDKI Jakarta Sebelum Swastanisasi (Condition ser-vice of drinking water in DKI Jakarta beforeprivatization) G. Pandjaitan (Technical DirectorPAM JAYA 1987–1991), Seminar ‘Drinking Waterin Jakarta’, organized by Poltak Simanjuntak (AKA-INDO) August, 2004, Jakarta, Indonesia.

PERPAMSI (1975a). Air Minum (Drinking water).WARTA Journal PERPAMSI, 1(1–2), 34.

PERPAMSI (1975b). Air Minum (Drinking water).WARTA Journal PERPAMSI, 1(3–4), 50.

PERPAMSI (1975c). Air Minum (Drinking water).WARTA Journal PERPAMSI, 1(5–6), 53.

Pierre, J. (Ed.) (2000). Debating governance. Oxford:Oxford University Press.

Pierre, J., & Peters, B. G. (2000). Governance, politics andthe state. Basingstoke: Macmillan.

Porter, R. (1996). The economics of water and waste: Acase study of Jakarta, Indonesia. Aldershot: Avebury.

Prasad, N. (2006). Privatisation results: Private sectorparticipation in water services after 15 years. Devel-opment Policy Review, 24(6), 669–692.

PSIRU (2006). Public–public partnerships as a catalystfor capacity building and institutional development:Lessons from Stockholm Vatten’s experience in theBaltic region. Public Services International ResearchUnit, University of Greenwich.

Renzetti, S., & Dupont, D. (2004). The performance ofmunicipal water supply utilities: Evidence on the roleof ownership. Journal of toxicology and environmen-tal health, Part A, 67(20–22), 1861–1878.

Rhodes, R. (1996). The new governance: Governingwithout government. Political Studies, 44(4),652–657.

Robbins, P. (2003). Transnational corporations and thediscourse of water privatization. Journal of Interna-tional Development, 15(8), 1073–1082.

Robeyns, I. (2005). The capability approach: A theoret-ical survey. Journal of Human Development 2005,6(1), 93–114.

Rogers, P., & Hall, A. (2003). Effective water gover-nance. Global Water Partnership Technical Com-mittee. TEC background papers, #7. GWP, Sweden.

Saleth, R., & Dinar, A. (1990). Evaluating waterinstitutions and water sector performance (Technicalpaper no. 447). Washington, DC: World Bank.

Saleth, R., & Dinar, A. (2000). Institutional changes inglobal water sector: Trends, patterns and implica-tions. Water Policy, 2, 175–199.

Saleth, R., & Dinar, A. (2005). Water institutionalreforms: Theory and practice. Water Policy, 7, 1–19.

Salim, P. (2005). Personal communication. PriyonoSalim, former Director for Technical Planning,Directorate of Urban Water Supply, Cipta Karya,Ministry of Public Works, Jakarta, Indonesia.

Sen, A. (1992). Inequality re-examined. Cambridge:Harvard University Press.

Sen, A. (1999). The capability approach: A theoreticalsurvey. New York: Knopf.

Server, O. (1996). Corruption: A major problem forurban management, some evidence from Indonesia.Habitat International, 1, 23–41.

Shirley, M. (2002). Thirsting for efficiency: The econom-ics and politics of urban water system reform. London:Elseveier.

Shofiani, N. E. (2003). Reconstruction of Indonesia’sdrinking water utilities: Assessment and stakeholdersperspective of private sector participation in thecapital province of Jakarta, Indonesia. TRITA-LWR. M.Sc. thesis. LWR-EX-03-30. ISSN 1651-064X. Department of Land and Water ResourcesManagement, Royal Institute of Technology, Stock-holm.

Silas, J. (1992). Government-community partnerships inKampung improvement programmes in Surabaya.Environment and Urbanization, 4(2), 35–36.

Simanjuntak, C. H., Larasati, W., Arjoso, S., Putri, M.,Lesmana, M., Oyofo, B. A., et al. (2001). Cholera inIndonesia in 1993–1999. American Journal of Trop-ical Medicine and Hygiene, 65(6), 788–797.

Strange, S. (1996). The retreat of the state: The diffusionof power in the world economy. Cambridge: Cam-bridge University Press.

Struyk, R. J., Hoffman, M. L., & Katsura, H. M. (1990).The market for shelter in Indonesian cities. Washing-ton, DC: The Urban Institute Press.

Surjadi, C. (2002). Public private partnerships and thepoor: Drinking water concessions case study –Jakarta, Indonesia. London: International Institutefor Environment and Development.

Surjadi, C. (2003). Public private partnerships and thepoor: Drinking water concessions, a study for betterunderstanding public–private partnerships and waterprovision in low-income settlements. Loughborough:Water, Engineering and Development Centre.

Surjadi, C., Padhmasutra, L., Wahyuningsih, D.,McGranahan, G., & Kjellen, M. (1994). Householdsenvironmental problems in Jakarta. Stockholm:Stockholm Environment Institute.

Susantono, B. (2001). Informal water services in metro-politan cities of developing world: The case ofJakarta, Indonesia, Berkeley. Doctoral thesis,Department of City and Regional Planning, Univer-sity of California.

Swyngedouw, E. (1997). Social power and the urbanisa-tion of water. Oxford: Oxford University Press.


Taylor, J. (1983). An evaluation of selected impacts ofJakarta’s Kampung Improvement Program. Unpub-lished doctoral thesis, Department of Urban Plan-ning, UCLA.

TNI (2005). Reclaiming public water: Achievements,struggles and visions from around the world. Amster-dam: Transnational Institute.

TNI (2006). Public water for all: The role of public–publicpartnerships. Amsterdam: Transnational Instituteand Corporate Europe Observatory.

Trawick, P. (2003). Against the privatization of water:An indigenous model for improving existing lawsand successfully governing the commons. WorldDevelopment, 31(6), 977–996.

Tutuko, K. (2005). Personal communication. Kris Tut-uko. Current Technical Director PAM JAYA,Jakarta, Indonesia.

UNCHS (2003). Local action for global goals: Water andsanitation in the world’s cities 2003. London: Earth-scan.

UNCHS (2006). Meeting development goals in smallurban centres: Water and sanitation in the world’scities 2006. London: Earthscan.

UNDP (2003). Millenium development goals: A compactfor nations to end human poverty. New York: UnitedNations Development Program.

UNDP (2006). Beyond scarcity: Power, poverty, and theglobal water crisis. Human development report 2006.United Nations Development Program, New York.

UNDP (2007). The global UNDP water governancestrategy: A follow-up to human development report2006: Beyond scarcity: Power, poverty and the globalwater crisis. United Nations Development Pro-gramme, New York.

UNWWAP (2003). Water for people, water for life.United Nations World Water Assessment Program.Rome: FAO.

UNWWAP (2006). Water: A shared responsibility.United Nations World Water Assessment Program.

van Breen, H. (1916). Beschouwingen: van de Techni-sche-en Watervoorzieningscommissies betreffende deverschillende in zake de voorgenomen verbeteringvan de watervoorziening der Gemeente Bataviaverschenen artikelen en uitgebrachte adviezen (Re-views: of the Technical and Water Supply commis-sions concerning the different articles and advicedealing with the intended improvement of waterprovision for the Municipality of Batavia) Gem-eenteraad van Batavia, Gedrukte Stukken No. 40(Municipal Council of Batavia, Printed Pieces No.40) (Ruygrok and Co., Batavia).

Van Leeuwen, C. A. E. (1917). Het huidige vraagstukder watervoorzieining van Batavia en Mr.Cornelis:voordracht, gehouden in de openbare vergaderingder afdeeling Batavia van 14 Februari 1917. DeWaterstaats Ingenieur, 3, 81–103.

Verhoest, K., Peters, B., Bouckaert, G., & Verschuere,B. (2004). The study of organizational autonomy: Aconceptual review. Public Administration and Devel-opment, 24, 101–118.

Wallsten, S., & Kosec, K. (2005). Public or privatedrinking water? The effects of ownership andbenchmark competition on US water system

regulatory compliance and household waterexpenditures. Brookings Institution working paper,05-05.

WaterAid (2003). Does PSP benefit the poor: Thesynthesis report. WaterAid and Tearfund, London.

Weimer, M. (2006). Action research on point of usedrinking water treatment alternatives as appropriatefor underprivileged households in Jakarta. Jakarta:Environmental Services Program.

Whittington, D. (1992). Possible adverse effects ofincreasing block water tariffs in developing countries.Economic Development and Cultural Change, 41(1),75–87.

WHO/UNICEF (2004). Meeting the MDG water andsanitation target. Geneva: WHO/UNICEF JointMonitoring Programme for Water Supply and San-itation.

WHO/UNICEF (2005). Water for Life: Making itHappen. Geneva: WHO/UNICEF Joint MonitoringProgramme for Water Supply and Sanitation.

WHO (1992). The international drinking water supply andsanitation decade: End of decade review. Geneva:World Health Organization.

WHO (2000). Global water supply and sanitationassessment. 2000 Report. Geneva: World HealthOrganization, UNICEF, and Water Supply & San-itation Collaborative Council.

Wilder, M., & Romero Lankao, P. (2006). Paradoxes ofdecentralization: Water reform and social implica-tions in Mexico. World Development, 34(11),1977–1995.

Winpenny, J. (1994). Managing water as an economicresource. London: Routledge.

Woodcock, J. (2005). Enabling water utilities to servethe urban poor. Jakarta: World Bank CountryOffice.

World Bank (1974). Appraisal of the Jakarta urbandevelopment project in Indonesia. Report No. 475a-IND, Transportation and Urban Projects Depart-ment, August 30.

World Bank (1991). Indonesia public expenditure forinfrastructure development. Washington: WorldBank.

World Bank (1993). Indonesia: Urban public infrastruc-ture services. East Asia Region, Jakarta: WorldBank.

World Bank (1994). Infrastructure for development.World development report, World Bank Group.

World Bank (1997a). The state in a changing world.World development report, World Bank Group.

World Bank (1997b). Indonesia urban water supplysector policy framework. Indonesia discussion paperseries. East Asia and Pacific Region Country Office,Jakarta.

World Bank (1998). Implementation completion report.Republic of Indonesia: Second JABOTABEK urbandevelopment project (Loan 3219-IND). Report Nr.18049. Washington, DC: World Bank.

World Bank (1999). Water sector adjustment loan:sectoral environmental assessment. E-267. Washing-ton: World Bank and Jakarta, Government ofIndonesia, Inter-Agency Task Force on Water SectorPolicy Reform.


World Bank (2004a). Making services work for poorpeople. World development report 2004. WorldBank Group, Washington.

World Bank (2004b). Indonesia: averting an infrastruc-ture crisis: A framework for policy and action.Indonesia Infrastructure Department. East Asiaand Pacific Region: World Bank.

Available online at www

Yayasan Dian Desa (1990). Monitoring and evaluationof public hydrants and water terminals in NorthJakarta. May 1990. Final report. Yogyakarta.

Zaidi, A. (2001). From lane to city: The impact of theOrangi Pilot Project’s low cost sanitation model.London: WaterAid.

.sciencedirect.com

Documents

Governance Failure: Rethinking the Institutional …...governance reform, and by raising the question of whether household provision of networked water supply by monopolistic providers