Technologies of Government: Constituting Subjectivities, Spaces, and Infrastructures in Colonial and Contemporary Jakarta


  • The research upon which this article is based was funded by a grant from the Social Sciences and Humanities Research Council of Canada. Assistance from research assistants Endah Shofiani and E.J. Martijn, and cooperation by water supply professionals in Jakarta and staff at local and international NGO offices during fieldwork are gratefully acknowledged, as is assistance from archival staff at colonial archives in Amsterdam and Leiden (the Netherlands) during research in 2003. The authors benefited from discussions with Abidin Kusno, Michael Leaf, Teti Argo and Joost Cote.

Michelle Kooy ( and Karen Bakker (, Department of Geography, University of British Columbia, Room 217, 1984 West Mall, Vancouver BC V6T 1Z2, Canada.



This article seeks to extend recent debates on urban infrastructure access by exploring the interrelationship between subjectivity, urban space and infrastructure. Specifically, it presents a case study of the development and differentiation of the urban water supply in Jakarta, Indonesia. Drawing on concepts of governmentality and materiality, it argues that the construction of difference through processes of segregation and exclusion enacted via colonial and contemporary ‘technologies of government’ has spatial, discursive and material dimensions. In particular, it seeks to ‘rematerialize’ discussions of (post-)colonial urban governmentality through insisting upon the importance of the contested and iterative interrelationship between discursive strategies, socio-economic agendas, identity formation and infrastructure creation. In exploring these claims with respect to Jakarta, the article draws on data derived from archival, interview and participant observation research to present a genealogy of the city's urban water supply system from its colonial origins to the present. We illustrate how discourses of modernity, hygiene and development are enrolled in the construction of urban subjects and the disposition of water supply infrastructure (and are also resisted), and document the relationship between the classification of urban residents, the differentiation of urban spaces and lack of access to services. The article closes with a discussion of the implications for analyses of the differentiation of urban services and urban space in cities in the global South.


Cet article tente d’élargir les récents débats sur l’accès aux infrastructures urbaines en explorant l’interrelation entre subjectivité, espace urbain et infrastructure. Plus précisément, il présente une étude de cas sur l’aménagement et la différenciation de l’approvisionnement en eau de Jakarta, en Indonésie. À partir des concepts de gouvernementalité et de matérialité, il fait valoir que la construction d’une différence par des processus de ségrégation et d’exclusion, mis en œuvre par des « technologies de gouvernement » coloniales et contemporaines, a des dimensions spatiales, discursives et physiques. Ce travail vise notamment à« rematérialiser » les discussions sur la gouvernementalité urbaine (post-)coloniale en insistant sur l’importance de l’interrelation contestée et itérative entre stratégies discursives, programmes socio-économiques, formation d’identité et création d’infrastructures. Tout en explorant ces idées dans le cadre de Jakarta, l’article exploite des données issues d’archives, d’entretiens et d’observations participantes afin de présenter une généalogie du réseau urbain de distribution d’eau, de ses origines coloniales jusqu’à nos jours. Il montre comment les discours sur la modernité, l’hygiène et l’aménagement s’inscrivent dans la représentation des sujets urbains et dans la disposition de l’infrastructure d’approvisionnement en eau (et comment s’exprime la résistance) ; de plus, il expose la relation entre la classification des résidents, la différenciation des espaces urbains et le manque d’accès aux services de la ville. La conclusion termine par les conséquences pour les analyses sur la différenciation des services urbains et de l’espace urbain dans les grandes villes des pays du Sud.


Jakarta's urban landscape is cluttered: spaces both beneath and above the city are traversed by a ‘tangled network’ of public, private and communal infrastructure for accessing, filtering and distributing a variety of water sources (Febrina, 2006). The majority of the city's residents rely on a complex mix of types of water and types of service provider: shallow and deep groundwater, piped network water, river water and bottled water on the one hand, delivered through pipes, pumps, private wells, ambulatory vendors, depots, tanks and networks on the other. Not surprisingly, less than one-third of Jakarta's estimated 12 million residents have access to a piped water supply in their homes.1 However, while all households use a heterogeneous mixture of supply sources, poor households typically access lower qualities and quantities of water, purchased at unit costs higher than those for the water available to wealthier households.

Over several decades the lack of access to the city's centralized piped network, and the associated impacts upon the urban poor, have provided the rationale for the initiation of numerous development programs with the ‘will to improve’ access to clean water in Jakarta (Bakker et al., 2006; Li, 2007).2 Premised upon analyses which treat the problems of persistent exclusion as apolitical and purely technical matters (see Li, 2007), successive programs of development have variously identified the lack of access to water supply infrastructure as a problem of corruption (Server, 1996), lack of finance (World Bank, 1974; Akhtar 2005; World Bank, 2005), or of rapid rates of urban growth outpacing infrastructural development (Hamer et al., 1986; Chifos and Suselo, 2000). Alternately, academic analyses critiquing these interventions point to the effects of neoliberal restructuring and its concomitant reshaping of urban infrastructure networks (Graham and Marvin, 2001). However, as this article documents, differential access to water supply has characterized the city since the colonial period, and has not been significantly altered either by programs of development or the rise of neoliberalism (see Bakker et al., 2006; Kooy and Bakker, forthcoming). How, then, are we to explain the persistent differentiation of access to water in Jakarta (and, indeed, in urban areas in the South more generally)? To answer this question, the article proposes an alternative conceptual framework which focuses on the relationship between governmentality, identity, urban space and urban infrastructure. We argue that the discursive construction of the ‘modern’ urban citizen that originally emerged in the colonial period, and which accompanied the introduction of a reticulated water network, was predicated upon and subsequently reinforced by racialized, and then class-based, divisions within the (post-)colonial city.

The core of the article presents a historical analysis based upon archival research3 in the Netherlands, and fieldwork4 conducted in Jakarta throughout the period 2004 to 2006. We begin, however, with a conceptual discussion of the interrelationship between subjectivities, spaces and infrastructures, before moving on to elucidate the links between the construction of ‘modern’, governable citizens, infrastructure and urban space in colonial Batavia and post-colonial Jakarta. Throughout, we focus on the interrelationship between urban subjectivities and urban infrastructures, arguing that access to different types of water supply infrastructure has been deployed both to define residents of the city as ‘modern’ or ‘in need of development’, and to rationalize exclusion from ‘modern’ water supply services. The article closes with some considerations of the relevance of this ‘rematerialization’ of post-colonial governmentality for the study of (post-)colonial cities and urban spaces. Countering post-colonial critiques of the overt focus on discursive relations of power (McEwan, 2003; Harris, 2004), we argue that the imbrication of infrastructure and identity is an important dimension to the ‘lived reality’ of residents of the post-colony.

Technologies of government: constituting subjectivities, spaces and infrastructures

Recent work on cities in the South has emphasized the interrelationship between urban identities and the spatiality of colonial and post-colonial power and discourse (see, for example, Bunnell, 2002; Kothari, 2006; Legg, 2006; Myers, 2006). Much of this work extends and critiques Foucauldian theories of governmentality. Paying specific attention to the process of constructing knowledges through which ‘subjects’ are governed, which they actively resist and reshape (Cooper and Stoler, 1989; Stoler, 1995; Li, 1999), post-colonial scholars also document how relations of power are enacted within physical space (Legg, 2006), and through material practices (Agrawal, 2005).

Some of this work has highlighted strategies of resistance by subjects, documenting nuanced negotiations between the ‘governed’ and ‘governing’ (Scott, 1985; 1992; Li, 1999; Young, 2001; Howell, 2004). Other scholars working within the analytic of governmentality outside of the Western liberal democratic state have emphasized the necessary contradictions and incompleteness inherent in government seeking ‘diverse finalities’ in the management of all of life (Li, 2007). This has been a useful means of extending analyses of governmentality to include the role of non-state actors (Rose-Redwood, 2006), and of producing post-colonial development geographies of urbanization which offer an alternative to the state-centric, often implicitly Northern-biased understandings of cities in the South (Power et al., 2006; Robinson, 2002; 2003).

This work also usefully illustrates the ways in which the construction of difference through processes of segregation and exclusion has both spatial and discursive dimensions. In this article, we attempt to broaden this approach to consider the material dimensions of post-colonial relations of rule. In particular, we contend that governmentality has a material as well as discursive dimension: relations of power are inscribed in physical space as well as social relations. Governmentality, in other words, has material effects; and material conditions play a role in constituting and/or contesting, government by constraining and shaping their form and effectiveness.

This conceptual stance draws in part on recent debates in political ecology, which begins from the assumption that ‘social worlds are as much constituted by materiality as the other way around’ (Miller, 1998: 3), and draws inspiration from critiques of the nature–culture binary in anthropology, sociology, geography and cognate disciplines (see, for example, Appadurai, 1986; Ingold, 1986; Descola and Pálssen, 1996; Thrift, 1996; Braun and Castree, 1998; Whatmore, 2002). Many applications of political ecology in the South have been to rural areas, and a significant subset have focused on environmental issues; Arun Agrawal's (2005) study of colonial governmentality, for example, demonstrates the mutual constitutiveness of colonial governmentality, the identities of colonial subjects and the environment (in this case, the forests of India's Kumaon region). But the mutual constitutiveness between what Agrawal terms ‘technologies of government’, subjectivity and materiality is, we would argue, equally applicable in urban settings. This assertion is, of course, not limited to urban areas; as recent work has shown, both rural and urban sites are (often mutually) implicated in post-colonial governmentalities (see, for example, Bebbington, 2000; Myers, 2006). But urban areas are of particular interest in part because of recent developments in urban political ecology and debates over ‘rematerializing’ urban geography, which have explored how urban infrastructure and technologies of government are mutually implicated in processes of urbanization (Swyngedouw, 1996; 1999; 2004; Gandy, 1997; 2002; 2005; Lees, 2002; Latham and McCormack, 2004; Heynen et al., 2006).

As Gandy and Swyngedouw in particular have shown with respect to colonial situations, the social relations of urbanization both shape, and are in turn shaped by, highly differentiated circuits of water flowing within the city (such as polluted rivers and canals for marginalized, migrant or ‘illegal’ urban populations, which overlap both as supply and effluent with the piped water supply for the affluent). As water supply infrastructure is long-lived (well over one hundred years in many instances), these circuits and networks simultaneously embody successive ‘relations of rule’ through the patterns of water supply infrastructure and water use practices they both enable and disable. An analytical framework of governmentality highlights how these relations are mobilized within particular systems of rule and exert specific effects upon the production of space and identity. In this case, we illustrate how relations around urban water are both a product, and productive, of a system of colonial authority that was based upon creating and reinforcing racial and class divisions. In other words, the construction of difference through processes of segregation and exclusion enacted via colonial ‘technologies of government’ has spatial, discursive and material dimensions — through the mutual constitution of subjectivities, urban spaces, and urban infrastructure. Elucidating these links is the focus of our case study, to which the article now turns.

Circulating ‘citizenship’: modernity, hygiene, and ‘rights’ to water

Producing native subjects: the colonial water supply system

As repeated epidemics of cholera and typhoid swept Batavia throughout the nineteenth century, demoting it from the ‘Queen’ to the ‘Graveyard’ of the East, water supply became a central concern for the Dutch colonial government (Abeyasekere, 1987). As elsewhere, new scientific theories of disease transmission via bacteria had begun supplanting formerly widespread notions of ‘miasmic’ contamination (Goubert, 1986; Gandy, 2006). Disseminated largely by military doctors in Batavia (Moens, 1873), these theories proved congruent with the Netherlands' evolving ambitions for its strategic colony. In the latter half of the nineteenth century, the Netherlands Indies colonial government began to legitimize its colonial authority through a racialized hierarchy of ‘development’ (Gouda, 1995). However, the existing blurred spectrum of native,5 Chinese, Arabic, Eurasian and full-blood European populations visibly contradicted, and complicated, the government's claim to racial superiority (Taylor, 1983). The ‘natural’ dominance of European over native (and Chinese and Arabic), and the new valorization of ‘modern European life’ as the colonial cultural developmental ideal was contradicted by an ‘Indische’ European population who integrated ‘native’ cultural practices into their households, and lacked clear biological markers of particular racial identities (Stoler, 1995). Hence, the production of differences to enable more visual distinctions to be made between racial categories was considered crucial to colonial control and the colonial government's campaigns against the ‘degeneration’ of European citizens into native life (see Stoler, 1995; Cote, 2003).

Batavia's first urban water supply exemplifies this broader project. Following a detailed scientific investigation into both the hygienic qualities of Batavia's waters, and into the (lax) hygiene behaviours of the city's ‘European’ residents (see Moens, 1873), the government exhorted the European residents to ‘take more care’ regarding their selection and consumption of drinking water, complementing the campaign with the provision of a clean water source through artesian hydrants. In the 1870s, the city's first centralized water supply system (using artesian water) was built. Intended only for the small minority of European residents living within a narrow strip of ‘development’ in the central area of the city (Van Raay, 1915a), the system provided free water to European areas of the city, but deliberately excluded ‘native’ residents. In contrast to the surface water from the city's kalis (canals) upon which natives relied, European households now had access to scientifically monitored groundwater, circulated through iron pipes rather than the corporeal networks of ambulatory water vendors.

Forty years after the system's construction, natives — already the majority of urban residents6— were gradually given access through a few public standpipes, but were not provided with access to household connections until the late 1920s (Maronier, 1929). Even then, records from the 1930s show that 90% of European residents were connected to the network, and used up 78% of the city's domestic supply, while comprising only 7% of the urban population.7 Native residents relying on mobile vendors rather than piped networks for their water paid more than double the cost to European residents (Heetjans, 1923).

Differential access to water supply infrastructure was mobilized discursively by the colonial administration to deepen the differences between European and native (or, increasingly, ‘modern’ and ‘primitive’) urban residents. The new water supply system enabled a specific set of material practices related to indoor hygiene that were coded as signifiers of a racially pure European identity.8 European households in the city were equipped with ‘every kind of modern convenience’— including ‘indoor bathrooms from which the fresh water from the tap may be showered over the body’ (Gemeente Batavia, 1937: 70) — which enabled water-intensive hygienic practices to be carried out daily, and in private. ‘Modern’ relationships between water supply, water quality and the human body were now defined through the scientific developments — notably in water quality assessments — made in the mid-nineteenth century9 (Moens, 1873). The colonial official who conducted the first extensive (and scientific) study of the various qualities of water supply in Batavia, for example, placed the water use practices of European residents under a microscope,10 arguing that traditional methods of water treatment —lekstonen and martevanen (a simple lime-sandstone filter and large water pots to allow the ‘settling’ of particulate matter in river water) — and quality surveillance (largely through individual assessments of colour, odour, and taste) undermined the morality and civilized status of the city's European residents (Moens, 1873: 416). Europeans were instructed to replace their traditional lekstonen and sensory assessments of water quality with scientific assessments of hygienic requirements (Maronier, 1929: 230) which were necessary to address the ‘unseen’ microscopic threats whose recognition was a marker of civilization (Moens, 1873: 292).

In contrast, practices such as bathing in canals and drawing drinking water from rivers could be, and were, denoted as ‘degenerate’ and unhygienic. Local scientific studies (largely carried out by military doctors) provided the basis upon which specific water sources were declared ‘sanitary’, and blame for cholera and typhoid epidemics was increasingly directed at natives' ‘hygienic circumstances’ (Maronier, 1929: 225). The identity of Europeans as modern and developed was secured through comparison with the native population. Colonial scientists noted, for example, that, ‘among the native population it is very common to drink murky water, not or not adequately filtered’ (Moens, 1873: 402), practices that had until recently also characterized ‘European’ households, but which they had now presumably ‘evolved’ out of.

The policing of these distinctions between European and native residents, never hermetic, proved a source of concern for the colonial administration (Stoler, 1992; Cote, 2003). Long-term residents and ‘degenerate Europeans’11 were problematized in colonial reports due to their ‘over-familiarity’ with native culture and practice of unacceptable ‘Indische’ traditions (Cote, 2003) — including preferences for ‘traditional’, untreated water sources and ‘unhygienic’ water practices. The problematic preference of some of the city's European residents for what were now considered ‘native’ water sources and bathing amenities was noted by the engineers recently arriving from the metropole. The slow growth of piped water supply from the artesian network was blamed on those who were ‘born in the Indies’ (i.e. racially mixed despite being legally European). Despite the extension of the artesian water supply system in European areas of the city, some residents did not ‘graduate’ to using the new artesian water source provided by the government.12

Somebody who has lived in Indië for years, or who was born there, lives in a house with a well in the inner courtyard, which is never the subject of complaints [over water quality]. From this well the ‘mandibak’ (manual-shower-basin) is filled and water is taken for further household life . . . The bathroom and toilet are, as with many older homes, somewhat backward in appearance and equipment, compared to current living standards (Koster, 1919: 5).

In contrast, ‘newer European homes’ built by the wave of middle-class, newly arrived, solidly ‘Dutch’ residents were ‘equipped with more hygienically designed bathrooms’ (Brandenburg, 1924) that were connected by a piped network to the artesian water hydrants, replicating the standards of ‘Dutch cities . . . [with] centralized water provision and other perfect technical and medical services’ (Koster, 1919: 5).

The production of difference between sources and qualities of urban waters, and ‘differently modern’ populations also facilitated the desired separation of urban spaces. The colonial government's construction of an artesian water supply system for European residents enabled the European population to distance itself from surface water both spatially (proximity of residences) and bio-physically (replacing drinking water sources with groundwater). With the urban landscape in a process of physical transformation from the decentralized riverside ‘garden villas’ of the ‘Indische’ or ‘Tempo Dulu’ colonial society to a more spatially segregated, and geographically concentrated, European population (see Milone, 1967), the construction of artesian hydrants within the growing clusters of European residences facilitated the respatialization of the city according to race, and provided contrasts between the ‘civilized’ urban life of the Europeans and the ‘primitive’ and rural lifestyles of native residents (see van der Kroef, 1954; Van Doorne, 1983). This project was so successful that by the 1920s visitors to the city observed that: ‘Batavia is a European town built by Europeans, except for the natives freely bathing in the canals and rivers’ (van der Kop, 1926: 149, emphasis added). As the development of the city's urban water supply infrastructure continued to parallel the patterns of European urban development in Batavia (Abeyasekere, 1989), the contrast in development and lifestyle between the European and native areas of the city reached a level which began to embarrass the colonial government. In the 1940s, the government's Visman Commission recorded the extremity of racial segregation in a city composed of ‘two different worlds’ (Van Doorne, 1983).

One marker of the difference between these two worlds was the disparate levels of water consumption which colonial authorities assumed were characteristic of different races. Remaining ‘rooted’ in their habits and traditions, and selecting water quality according to colour, taste and smell, native residents were not as ‘modern’ or developed, and so were considered by colonial authorities to have ‘less need’ for access to the city's piped water supply infrastructure. An engineer (and former Officer of the Royal Engineers in the Netherlands Indies) appointed by the Municipal Council to review the designs for the city's subsequent water supply system13 calculated the water needs of the city predicated upon the different water use habits of distinct racial groups: in Van Raay's design the total demand was based upon the prediction that 100% of Europeans would use 150 litres per capita per day, 60% of the Chinese/Arabs would use 100 litres per capita per day, and 20% of the native population would use 50 litres per capita per day (Van Raay, 1915b; van Leeuwen, 1917). ‘Non-Europeans’ would, it was accepted, continue to meet their water needs in other ways:

if one asks about the different categories, why widely varying usage figures are assumed [between races], then the reason is partly to be seen in the fact that, per lifestyle (mode of living), some have less water needs than others . . . [for natives] a large part of the water used for internal purposes [human consumption] is taken from the water supply network, while bath and wash water is taken from the existing bad wells . . . [so] we know that an average of 50 litres per day is not enough for the native, this is used as a starting point for the design of the water system (Gomperts, 1916: 3).

Perversely, while the limited extension of the piped water supply into native areas of the city was rationalized by their lack of ‘development’, the natives' limited ability to access (and afford) this water supply reinforced their status as ‘undeveloped’. Continuing to use river water for the majority of their needs, native residents thereby confirmed their ‘primitive’ status, as they did not observe the delineation of proper, private spaces for domestic water use. The kalis were of particular concern. Batavia's canals were ‘scenes of considerable activity from early morning till late afternoon. The banks of the canals, the steps that lead down into them, and the bamboo rafts are all crowded with throngs of half-naked native washerwomen and laundry men’ (Vervoort, 1926: 266). Colonial hygiene officers and engineers considered this to be a public display of ‘their undeveloped approach to the functions of life’ and their ignorance of the ‘proper forms of urban life’ (Van Breen, 1919: 138). Observing native residents using the city's canals and rivers for bathing and laundry, colonial chroniclers described the natives ‘glimpsed in Batavia's canals’ as ‘children’ in both their practices, and mindset (Vervoort, 1926: 266). Conveniently for the colonial government, this justified the continued need for colonial control; the ‘children’ of the Netherlands East Indies would continue to require the guidance of their colonial parent.

The continued threat to public health and morality posed by large-scale use of untreated groundwater and surface water sources for drinking water and sanitation within the city resulted in repeated calls to extend the water supply network to kampongs throughout the duration of the colonial administration. The ‘deeply rooted evil’ of public bathing (Van Breen, 1919: 131) was, for example, targeted by the Municipal Council though the construction of communal, yet enclosed, washing and bathing facilities for native residents — to ‘keep the [native] population out of the canals’ (Van Raay, 1915b: 142). These facilities established ‘private’ spaces for domestic functions, and simultaneously differentiated the waters in which it was ‘proper’ to wash oneself, one's clothes and one's household items. Native residents had to be led to rationalize their use of urban space (and water) according to what was ‘hygienic’ and proper; bathing in open waterways was ‘distasteful’, and the ‘dignified [i.e. educated] adult native’ was to be taught the proper ways to divide bathing and cleaning from recreation, and the proper spaces and sources of water for these newly distinct functions (Karsten, 1958: 42).

Yet projects to ‘improve’ native hygiene were riven with contradictions, and surprisingly (at least to the colonial authorities) met with resistance. Resistance surfaced through vandalism of standpipes, and a widespread refusal of natives to pay for water from the public standpipes that were gradually extended into the kampongs (Maronier, 1929; Argo, 1999). This, in turn, stymied further extension of the network, intended by the colonial authorities to be run on a cost-recovery basis (unlike the free connections available to Europeans). Yet the failure to extend the network provoked reactions from the ‘modern natives’ who had begun to emerge out of Batavia's ‘cosmopolitan’ urban environment. This new type of colonial subject embraced Western technologies in order to criticize the ways in which they were used to perpetuate, rather than eliminate, the inequities of colonial rule (Kusno, 2003). For example, the new spring water network created in Batavia in the 1920s was mobilized as a symbol in anti-colonial criticism. Pointing to the ‘drifting of colonial action on calls to improve water supply and sanitation in native communities’, modern native subjects complained about the lack of water supply infrastructure in native areas of the city, and argued for a more equitable distribution of clean water to reduce native mortality rates and disease (Mrazek, 2002).

The evolution of ‘natives’ into modern residents threatened to disrupt the colonial system of classification, and their appropriation of the issue of the piped water supply for political purposes was perceived as a threat to the established colonial order. When the ‘newly modern’ native residents criticized the contradictory ways in which the government engaged the colonized with modernity, the differences in the physical and spatial nature of the city's ostensibly ‘public’ water supply became mobilized as a critique of colonial government. The ‘modernization’ of natives was, from this perspective, regrettable:

Many of those who have been longer in the East than just a few years refer with a certain disdain to the native population in and near the large European towns, and especially many government officials and planters are always prepared to extol the virtues of the village native in the interior, the simple peasant class, in contrast with the town native, who has come into touch with European civilization. (Van der Kop, 1926: 153).

Natives were not meant to be modern — or at least not for another century. Remaining visibly ‘rural’ in lifestyle, and primitive in habit, the natives ‘bathing in the city's canals’ were at once both ‘distasteful’ and necessary to colonial government. Pictured in an overwhelming number of the colonial documents on urban development, the ‘natives in the canals’ both enacted and embodied the contradictions inherent within a colonial government that mobilized different modalities of power (disciplinary, sovereign, government) upon differently positioned subjects (rulers vs. ruled) for ‘diverse finalities’.14

Producing modern subjects: the post-colonial water supply system

The image of the ‘native in the canals’ continued to provoke and contradict governmental intentions following Independence in 1949. President Sukarno's global ambitions as a leader in the non-aligned movement (most notably via the Bandung conference in 1955) and his national ambitions of developing and unifying a multi-ethnic, multilingual and shaky nation converged, in part, in a personal project to modernize and beautify Jakarta (Kusno, 1997). However, the public monuments and highly visible infrastructure projects initiated under Sukarno's direct oversight were visibly undermined by the ‘masses of poor’ residents and their ‘public striptease’ of washing and bathing in the city's canals, compromising its identity as a ‘beacon for the third world’ (Hanna, 1959; Kusno, 1997). Residents with ‘backward urban lifestyles’ who did not contribute to making Djakarta a ‘Beautiful, Orderly, and Dynamic city’ by adopting the ‘good norms and standards of city life’ (DKI, 1972: 85) were displaced to the non-serviced periphery of the ‘international’ area of the city.

At the end of the first decade of independence a piped water supply was only available to 12% of Jakarta's population (Fischer, 1959) — and this was considered an achievement given previous years of even lower coverage (Hanna, 1959), and the city's exponential population growth after independence.15 In the 1960s, it was estimated that only 15% of the city's residents were served with a household connection (Pam Jaya, 1992), and as the city continued to grow in both area and population, the majority of the city's residents remained excluded from service. The informal, low-income ‘kampong’ areas absorbed the majority of low-income migrants to Jakarta,16 and 90% of kampong residents did not have access to piped water supply (KIP, 1976). By the late 1980s the provision of piped water still only extended to less than one-quarter of the city's population (Porter, 1996), and in the mid-1990s, only 10% of lower-income residents were directly connected, while over 60% of the upper-income population had direct access to the network water supply (Cestti et al., 1994; Porter, 1996; JICA, 1997; Azdan, 2001). Large tracts of the poorest areas of the city are without access to water supply networks altogether (Forkami, 2006; Sabarini, 2007; Kooy and Bakker, forthcoming).

The stark disparity in water access echoes the spatial and social distance between the European and native populations in the colonial period. While the upper class in Jakarta import Western bathtubs, whirlpools and home spa sets to ‘turn the bathroom into a relaxing retreat’ (Lubis, 2004), the most impoverished residents of the city scavenge wood scraps to construct ‘helicopter toilets’ over rivers and canals to provide themselves with a bare minimum of privacy and ‘comfort’. Preferring to soak in a bathtub or take a shower rather than use the ‘traditional’mandi-bak, the urban elite are now ‘too modern’ for the level of infrastructure that most urban poor have not even yet achieved. The dramatic difference in urban lifestyles and infrastructure illustrated by this conspicuous consumption of Western lifestyle products is justified as the pursuit of a ‘modern lifestyle that emphasizes efficiency, effectiveness, comfort and healthy living’ (Wiradji, 2004: 1).

Perversely, as in the colonial city, the ‘consistent exclusion’ of kampongs from the city's ‘public’ water supply system has left residents with little choice other than to perpetuate their ‘unmodern’ identities.17 Living in the unserved interstices of the ‘modern’ city, low-income residents build their own shallow groundwater wells and consume dubious-quality water, or pay up to 15–33 times the official water tariff for a vended water supply (Fidrus, 2006; Mukherjee, 2006). With 80% of the city's shallow groundwater, and almost 100% of its surface waters, contaminated by e-coli and/or heavy metals (Harsanto, 2005; Simamora, 2007), the persistent use of these free sources of water for washing and bathing (and sometimes drinking) by the urban poor enables their classification as ‘backward’, not scientifically rational enough to accept the health risks within a water source that they cannot see, smell or taste. Ironically, however, at least one medical study has indicated that e-coli contamination is lower in shallow groundwater than in the piped water network (which does not supply potable water, and is subject to frequent pressure reductions and leaks which compromise water quality, most particularly in low-income areas) (Surjadi et al., 1994). Dependent upon local, context-specific, experiential information (Scott's ‘metis’) to determine the safety of different sources of drinking water, residents' perceptions of clean and safe water are associated with an absence of odour (including residual chlorine) and sediment; even those with access to the city's piped network water supply system will aerate their drinking water in open buckets prior to consumption to dissipate the residual chlorine smell from treatment (Aman Tirta/USAID, 2006).

As in the colonial period, contemporary development problematizes ‘traditional’ sensory assessments of water quality as well as the public use of surface waters for domestic functions by the urban poor. For example, while justifying its interventions into hygiene behaviours along with infrastructure investment in Jakarta,18 USAID's Formative Research records that the idea of ‘cleanliness’ amongst the poorer communities is ‘based on culture and tradition’, rather than bacteriological criteria (ESP, 2006). Illustrating the need to develop the people along with the pipes, focus group discussion reports tell us that ‘mystical’ belief systems still inform people's perception of sickness, disease transmission, and influence patterns of water use from ‘sacred springs’ (ibid.). Similarly, the World Bank's ‘Voices of the Poor’ (Mukherjee, 2006) uses statements about water quality and preferences from the poor to frame them as objects in need of development. One site visit documents that, ‘Almost all people wash their clothes, take a bath and defecate at the river even though they have a well. Defecation in the river is perceived as “clean”, as it does not create a bad smell’ (Mukherjee, 2006: 26).

While the stated preferences of respondents for defecating in running water and their related perception of cleanliness is used to display ignorance and need for education in hygiene behaviour to correct the ingrained habit of ‘the strong cultural preference’ (ibid.), the discourse used to interpret the ‘voices of the poor’ echoes the colonial past, when the ‘natives taking pleasure bathing, washing and defecating in streaming water’ was demonstrative of ‘their insensitivity to cleanliness and order’ (Van Leeuwen, 1920: 198). Preferences for local water sources, ‘traditional’ assessments of water quality, and the use of public spaces for domestic functions are discursively situated in the cultural past — as the residents of Jakarta who use ‘polluted’ water supplies and combine a variety of different quality water sources as a livelihood strategy are not presented as being in need of ‘access’ and a policy change (such as tenure reform to enable legal water supply connections for squatter households), but rather as being themselves in need of cultural and value change (see Mujianto, 2004; Mukherjee, 2006).

Ironically, without access to more sophisticated technologies for assessing water quality and more reliable water sources, these urban residents will continue to use locally available, easily accessible and free water sources. Indeed, the choice to do so offers important, symbolic vehicles of resistance to the urban population, often distrustful of state authority that all too often has forced on them forms of ‘development’ that do not make ‘sense’ according to their standards of living (Jellinek, 1991; 1997). Using the public fountains in the city's iconic Independence Square (Monas) as a place for bathing and laundry (Harsanto, 2005), low-income residents have supplanted the intended ‘spiritual symbolism’ of this urban space (see Kusno, 2000) with politically symbolic calls for the government to meet immediate material needs. Illegal connections (estimated at 40,000) have provided another important means of resistance (Harsanto and Wahyudi, 2002). As in the colonial past where the sudden imposition of a system of ‘paid kampong water supply’ provoked resentful residents to tamper with pipelines and meters (Maronier, 1929), the urban poor who are ineligible for legal connections pay higher rates to informal ‘middlemen’ (often employees of the municipal water supply company) to provide them with services. Currently, 50% of Jakarta's network water supply is lost through ‘unaccounted for water’ (Nurbianto, 2007), 80% of which is estimated to be from ‘administrative’ rather than physical leakage (Castalia, 2006; Sabarini, 2007; and a personal communication from Forkami in 2007). And, rather than change their perceptions of what constitutes ‘clean water’, low-income residents are opting for another surprising — yet cost-effective — alternative that complicates their portrayal in development discourses. Scavenging garbage dumps for the empty 19 litre water bottle containers which can be ‘refilled’ at the ‘air isi ulang’ (water refill) depots, the urban poor are now — like the rich — using bottled water for drinking.19


In colonial Batavia and post-independence Jakarta, the project of producing ‘civilized’ or ‘modern’, developed citizens to ‘properly’ inhabit a productive, hygienic urban environment entailed not only discursive reworkings of the rationalities supporting water supply delivery, but also physical reworkings of urban space, and subjective repositionings of urban identities. Residents whose domestic water practices did not demonstrate a familiarity with scientific rationalities, modern concern for bodily health, or an appropriately economical use of water have been marginalized both materially and discursively. Access to a ‘modern’ piped water supply has continued to remain contingent upon one's identity as a ‘modern’ citizen — signified in part by a scientific understanding of water quality, the possession of privatized spaces of water use, and an economically ‘rational’ understanding of water costs/benefits.

As the colonial and post-colonial governments pursued projects of ‘civilizing’ or ‘developing’ their subjects, the associated classification of categories of urban citizens was translated into differentiated urban spaces; this was both rationalized and reiterated through the urban water supply infrastructure. In other words, the production — and selective distribution — of a ‘modern’ urban water supply has physically facilitated and reinforced the initial discursive (and political) division of populations and urban spaces. Physical and discursive divisions between populations, spaces and waters became entrenched within the urban landscape with the creation of the colonial water supply infrastructure. Post-independence discourses of modernization, hygiene, and water supply further cemented these divisions, as network expansion was confined to areas of the city inhabited by ‘modern’ and ‘developed’ urban residents. Although the discursive emblems associated with water have largely shifted from racial to socio-economic markers of status, through the (post-)colonial era, discourses linking water use with identity have remained central to determining informal rights to access different categories of public water supply services and physical access to such services.

This analysis has been predicated upon (at least) two claims that are of broader relevance to current debates over post-colonialism and urbanization. First, discursive binaries enrolled in the projects of colonization or modernization — European/native, modern/primitive, civilized/uncivilized, sanitary/insalubrious, developed/backward — are central to the technologies of government enacted in both colonial and post-colonial contexts, but documentation of the discourses is not sufficient to explain the persistent exclusion of the majority of Jakarta's residents from access to the water supply system. In ‘materializing’ the analysis of Batavia/Jakarta's technologies of government, this article has insisted upon the importance of the interrelationship between discursive strategies, socio-economic agendas, identity formation, and infrastructure creation; an iterative process, with changing patterns of socio-spatial access to water supply as an artefact. The material practices and urban spaces within and for which one uses urban water continue to indicate one's level of hygiene, cleanliness, and suitability for ‘modern’ urban life in Jakarta — and have been used to demonstrate one's ‘right’ to the use of modern urban infrastructure, as well as shaping decisions about the extension (or withholding) of infrastructure to specific spaces within the city.

Second, resistance is central to the constitution of these technologies of government, and has its own material effects. The persistence of preferences for local and easily available, yet poorer-quality, water sources by the urban poor is usually interpreted as their ‘failure’ to develop, but it can also be seen as resistance to a pattern of water supply infrastructure development that has consistently made it more difficult for the poor to access clean water. Acknowledging excluded groups as active agents in the constitution of such technologies of government is a necessary conceptual dimension of the rematerialization of urban governmentality. Therefore, although the spatial and material effects of government have reinforced their discursive positioning of Jakarta's less wealthy residents as ‘undeveloped’, the continued practice of behaviours labelled as ‘primitive’ can also be viewed as the material product of a strategy of active, informed resistance to urban governance policies that have sought to create the capital city as one large ‘gated community’— denying entrance, livelihoods and citizenship to those who are ‘not modern’,20 and who do not (or cannot) conform to the model of ideal citizens.

To conclude, it is not our aim to romanticize the current situation amongst urban poor populations in Jakarta; the health and related economic implications of relying on poor-quality water are strong reasons to pursue projects of more equitable access to clean water supplies. Rather, we have argued that more equitable approaches to urban infrastructural services require a historicized and politicized understanding of the cultural and discursive as well as economic and technical dimensions of exclusion. Lack of access to infrastructural services in cities in the global South is predicated upon the mutual reinforcement of material, spatial, and discursive aspects of ‘technologies of government’ reliant upon exclusionary models of citizenship. This suggests the need not only to take seriously the discursive relations of power enacted by colonial governments, but also to pay close attention to how these relations have been materialized in identities, urban spaces, and infrastructures in contemporary cities of the Global South.


  • 1

    The official (generous) estimate made by the Jakarta Water Supply Regulatory Body in 2004 was that 56% of the city's residents were connected to the network, but this figure included residents who accessed the network through public standpipes and mobile water vendors. Also, coverage ratios should be understood as rough estimates; their calculation is dependent upon a number of variables which are only imprecisely measured, such as urban population and average size of household. Reported figures vary significantly, and do not indicate the number of households which have a connection but which rely primarily on other sources (e.g. groundwater) due to quality or service concerns (e.g. low pressure).

  • 2

    The term ‘will to improve’ is borrowed from Li's (2007) chronology of (post-)colonial governmentality in Indonesia.

  • 3

    The Nationaal Archief in Den Haag, KITLV (Koninklijk Instituut voor Taal-, Land- en Volkenkunde) in Leiden, and KIT (Koninglijk Instituut Tropical) in Amsterdam.

  • 4

    Fieldwork entailed: participant observation with local and international NGOs involved in development and/or activist programs in urban poor communities; interviewing local and international water supply professionals in the private sector, donor and engineering community; and discourse analysis of various media, development donor and activist documentations of water supply in Jakarta from the 1950s to 2006.

  • 5

    It is important to note that ‘native’ was by no means a homogenous category, it comprised a wide range of ethnic groups from across the archipelago: Betawi, Buginese, Madurese, Balinese, Sundanese, Javanese and more (see Castles, 1967).

  • 6

    In 1875, three years after initial construction of the artesian water supply, native residents made up four-fifths of the urban population (Abeyasekere, 1987).

  • 7

    In 1929, 6,926 kampung households were supplied with 24 litres, while 10,392 European households were supplied with 84 litres. The European population in Batavia in 1930 was 37,067, while the native Indonesian population was over 400,000 (Eggink, 1930).

  • 8

    Stoler (1995) emphasizes the constitution of European identity through ‘culturally coded practices’.

  • 9

    The discovery of bacteria and the development of the science of hygiene established new relations between the water supply and human health, see Goubert (1986) and Melosi (2000).

  • 10

    Colonial scientists conducted laboratory tests on various sources of ground water, and surface water, giving readers the detailed biochemical content of each source, and including an illustrated appendix on diagrams of the microscopic organisms (Moens, 1873).

  • 11

    The ‘degenerate’ European residents were largely of mixed descent; legally considered as ‘European’, these residents were the offspring of the ‘long time colonial settlers’ who had adopted an Asianized urban culture, see Taylor (1983) and Milone (1967).

  • 12

    In 1923 a housing ordinance addressed this problem, making connection to the city's piped water network mandatory for all houses of a certain socio-economic class (monthly rent over 25 guilders), the vast majority of which were occupied by European residents (Eggink, 1930).

  • 13

    A spring water supply system delivering pressurized, potable water into the city was completed in 1922 (Smitt, 1922).

  • 14

    Li (2007) describes some of the ‘diverse finalities’ of colonial government in the Netherlands East Indies as securing orderly rule, entrepreneurial profit and native improvement — all of which were in tension with each other.

  • 15

    In 1948 prior to official independence, Jakarta's population was 823,000; by 1952 the population was estimated at 1.8 million, and by 1965 this was 3.8 million (Cowherd, 2002).

  • 16

    In the 1970s, the informal, low-income communities (kampongs) contained 80% of Jakarta's population and made up 65% of its land area (World Bank, 1974).

  • 17

    The history of the development of Jakarta's water supply infrastructure, and documentation of persistent exclusion of particular urban populations is provided in Bakker et al. (2006) and Bakker et al. (forthcoming).

  • 18

    USAID funding to increase access to basic services across Indonesia allocated US $311 million from 2004–8. The sectors of health, nutrition, safe water supply and sanitation are covered by the programs: Basic Human Services (BHS), Health Services Program (HSP) and Environmental Services Program (ESP) (http://www.indoensia/

  • 19

    Supporting a huge growth in the number of small-scale businesses selling generic bottled water, these refill stations provide treated, potable water at less than one-third the price of the brand names. Generic bottled water from the ‘air isi ulang’ (refill stations) is 3,000 Rupiah for 19 litres, whereas the Aqua brand preferred by the middle-class sells for approximately 10,000 Rupiah for 19 litres.

  • 20

    The ‘closed city’ policy of the city government in the 1970s failed after a few years, but current ‘police raids’ on migrants and unregistered urban populations still continue to ‘crack down’ on those who do not fit the socio-economic profile of ‘desirable’ residents of Jakarta (Harsanto, 2004; Fidrus, 2006).