(In)visible urban water networks: the politics of non-payment in Manila’s low-income communities

I. Introduction

At the Asian Development Bank’s (ADB) 2010 conference on water, the first breakout session of the five-day event was devoted to increasing utility efficiency, specifically through the reduction of non-revenue water (NRW).⁽²⁾ NRW is the percentage of water that enters the distribution system but is not billed to customers, and a reduction in NRW, the panelists claimed, could be seen as a critical step towards improved utility performance. One panelist stated that it was a relatively new topic of discussion, having been recognized as an important issue only in the last decade, while another claimed that it now made much economic sense because reducing NRW was significantly less expensive than new source development. The statistics that the panelists provided were impressive: Phnom Penh’s NRW fell from 72 per cent in 1993 to an astonishing six per cent in 2006; Hai Phong’s NRW, down from 70 per cent in 1993 to 25 per cent in 2008; and Manila Water’s NRW, from 63 per cent in 1997 to 16 per cent in 2009, and still decreasing.

These statistics may be even more impressive given their context – all three utilities are situated in large, metropolitan areas of the Global South, where prior expansionary efforts have often been piecemeal and incomplete, taking place under the aegis of different authorities and administrations. In these cities, formal infrastructure tends to be hidden underground,⁽³⁾ while informal infrastructure may consist of interwoven “spaghetti networks” that are likely unregistered and unmapped.⁽⁴⁾ The challenge of improving utility efficiency in developing cities often involves a reconfiguring of networks so that they become more visible to those seeking to upgrade them. This amounts to a significantly greater task than that faced by cities with well-established infrastructure, where maps and other data provide a better framework for understanding underground networks.

This paper describes processes aimed at increasing the visibility of urban water networks in Manila, the site of the world’s largest water privatization project.⁽⁵⁾ It focuses on the two private concessionaires’ efforts to improve efficiencies by reducing NRW – examining, in particular, commercial losses that are caused by theft and non-payment, meter under-registration or data errors.⁽⁶⁾ Unlike physical losses, which are leakages that occur throughout a utility’s system and are largely technical problems, losses due to theft and non-payment are necessarily social and political issues. The manner in which they are addressed and deterred thus provides some insight into the nature of urban water governance and the implications of these new governance initiatives for citizens. Non-payment, as discussed below, is a logic that binds both poor and non-poor customers, but it is the concessionaires’ differentiated treatment of these two groups that bears further examination. These policies are explored in the remainder of the paper, using the impact of NRW reduction strategies on low-income households as a lens for re-thinking strategies for basic needs provision.

II. Manila’s Water Privatization Project

The greater Metro Manila area is a sprawling urban agglomeration composed of 16 cities and one municipality, with an officially recorded population approaching 12 million.⁽⁷⁾ By the mid-1990s, the state-run Metropolitan Waterworks and Sewerage System (MWSS) was faring poorly compared to its Asian neighbours, supplying water to only two-thirds of its intended recipients for an average of 16 hours a day, with NRW at 56 per cent.⁽⁸⁾ But while the system had been suffering for decades, two historical moments were instrumental in the 1997 privatization of MWSS. The first involved the dramatic increase in private sector participation (PSP) in the water and sanitation sector in the 1990s – a shift promoted heavily in the Global South by institutions such as the World Bank and its investment arm, the International Finance Corporation (IFC).⁽⁹⁾ PSP was seen as a means of improving utility efficiency and self-sufficiency as well as supplying capital investment for poorly financed public systems. But although the IFC oversaw the MWSS privatization project, another historical moment must also be considered – the Philippines’ homegrown neoliberal turn, which was influenced by both global neoliberal shifts and national responses to the aftermath of Marcos-era cronyism.⁽¹⁰⁾ Under President Ramos, the Philippines witnessed the privatization of a failing power generation sector and the state-owned oil company. A severe, El Niño-related drought led to Ramos’ enactment of the National Water Crisis Act of 1995, which empowered the president to privatize the water sector. The privatization of MWSS was thus the result of a confluence of ideological beliefs originating from the state and international lenders, with relatively little resistance.

Per IFC’s advice, the MWSS service area was halved, in the style of the “Paris model”, and 25-year concessions were awarded to Manila Water on the east side (which began as a partnership between International Water, composed of United Utilities and the Bechtel Corporation, and the Ayala Corporation) and Maynilad on the west side (originally a joint venture between Lyonnaise des Eaux and Benpres Holdings).⁽¹¹⁾ The challenges of expanding coverage included ageing, unknown infrastructure – particularly on the western side, where the oldest portions of the metropolitan area and water system lie – as well as high levels of theft and non-payment among both the poor and non-poor. Furthermore, the 1997 Asian financial crisis proved to be particularly debilitating for Maynilad, which had been saddled with 90 per cent of MWSS’ debt as part of the concession agreement. As a result of these circumstances, Maynilad’s original owners returned the concession to MWSS in 2002, and it was not until 2006 that new investors took over.⁽¹²⁾

Despite these hurdles, Manila Water appears to have performed exceedingly well, and even Maynilad’s track record is improving. In 2010, Manila Water reported that 99 per cent of the population in its jurisdiction had 24-hour water availability, while NRW was down to 11 per cent; Maynilad claimed 71 per cent coverage and 53 per cent NRW.⁽¹³⁾ Furthermore, Manila Water’s flagship programme, Tubig Para Sa Barangay (“Water for the poor”, henceforth referred to as TPSB), has garnered numerous national and international awards for its delivery of services to low-income communities, including informal settlements. Based on these accounts, it appears that Manila Water has been able to address many of the critiques of privatized water delivery – such as failure to serve the poor – that have been made of utilities in Buenos Aires, La Paz and Cartagena, among others.⁽¹⁴⁾

But while Manila Water (and, to a lesser extent, Maynilad) has made vast improvements to the system, this paper focuses on the limitations of those improvements among low-income communities and households. As mentioned, the success of water utilities is largely evaluated through aggregate metrics such as coverage, NRW and other measures of efficiency, despite their self-reported and therefore biased nature. The ADB report on NRW concedes: “Most of the published water utility data, especially from low- and middle-income countries, are unaudited, resulting in sometimes significantly incorrect data being included in databases and publications, which can be very misleading.”⁽¹⁵⁾ Yet, for lack of better options, it is this data that ADB and other international organizations use to benchmark utilities across regions. And it is arguably the pressure exerted by organizations and investors to become “model utilities” that drives utilities to downplay how their services may be lacking.

III. Addressing Non-Payment

Those who engage in non-payment do so for a variety of reasons and – as evidenced by pre-privatization NRW estimates – may do so with some significant frequency and, over time, by significant amounts. Faced with the challenge of reducing NRW, the two Manila concessionaires have taken measures that target both poor and non-poor alike. These include the following:⁽¹⁶⁾

While the urban poor are often assumed to be the main culprits of water theft, the concessionaires and international organizations generally acknowledge that non-payment among the poor constitutes a relatively small percentage of the commercial losses.⁽¹⁹⁾ That is, even if there were many non-paying households among the urban poor, the volume of water that they each consume – as well as the resulting lost revenue – is relatively low. Rather, the bulk of commercial losses and missed revenues have come from industrial establishments and other large users. In 2000, for instance, Maynilad revealed that the Coca-Cola Bottling Company had an illegal connection installed in 1984 that accounted for PHP 27 million of arrears.⁽²⁰⁾ Recognizing that a relatively small percentage of consumers are responsible for the bulk of water revenues as well as non-payment, several specific measures have targeted these high-volume non-payers:

IV. Deterring Non-Payment In Low-Income Communities

As the concessionaires have expanded their operations into low-income communities, they have modified the terms for accessing water in order to deter non-payment. The result is a partial transfer in monitoring responsibilities – from the utilities to individuals and communities. Such a shift can be seen as a means of depoliticizing NRW for the concessionaires, as they take on a greater technical role in NRW management while reducing some sociopolitical engagement. But rather than framing this shift as such, the concessionaires couch access to water in the language of partnerships, ownership and morality – referring, often, to a sense of “Filipino values”. Regarding new connections provided to low-income residents, a Manila Water manager remarked: “The Filipino value of utang na loob (debt of gratitude) comes in… community policing comes in because they have a connection… they have a sense of ownership in the project so they themselves watch their meters.”⁽²⁵⁾ International organizations have also helped to paint a similarly rosy picture. A United Nations report describes Manila Water’s TPSB programme as fostering partnerships that “…enhance the community’s sense of ownership and increase the willingness to pay, encourage residents to closely monitor and guard against pilferage, improve collection efficiency, increase transparency and expedite public consultations, all of which make the TPSB programme manageable, financially viable and sustainable.”⁽²⁶⁾ Although these characterizations may very well have some truth to them, they also mask the underlying reasons behind this newfound sense of ownership.

Rather, as critiques of similar partnerships in other cities have demonstrated, these relationships can not only effectively transfer costs and responsibilities from utilities to low-income households but also paternalistically demand behavioural changes. Writing about various partnership schemes in sub-Saharan Africa, Jaglin demonstrates how such models have been used to offload some utility management risks onto low-income communities, describing this externalization as the “miracle of participation”.⁽²⁷⁾ Likewise, von Schnitzler examines South African attempts to “…empower customers to take ownership of consumption…” through the forced calculability of pre-paid meters.⁽²⁸⁾ In Mumbai, McFarlane considers attempts to use participatory slum sanitation schemes to “…foster a particular kind of civic consciousness of community responsibility deemed lacking among those living in informal settlements.”⁽²⁹⁾

This section examines two models that have been used to provide “more accessible” water connections to low-income communities in Manila – micro networks and clustered meters. The scale of these two models has been very difficult to estimate.⁽³⁰⁾ One statistic reflects that Manila Water once operated nearly 800 micro networks, although more than half had been converted by 2010 due to pressure from the MWSS Regulatory Office to individualize connections.⁽³¹⁾ Meanwhile, Maynilad has served more than 2,000 households through its Samahang Tubig Maynilad programme, which delegates water management to community-based organizations, and it projected that another 2,000 families would be served in a similar manner in the near future.⁽³²⁾ Clustered metering has only been used on Manila Water’s side, but it appears to comprise a significant portion of the TPSB programme, which had served 1.7 million people by 2011.⁽³³⁾ It is important to note that these are not universal policies and that there are low-income households that have more equitable terms for accessing water. As a general rule, households with proper land titles and that are not situated in narrow alleyways may be eligible for individual connections; in the standard version, a meter is placed immediately outside the house. However, the purpose of highlighting these two scenarios is to illustrate the limits of water provision where financial goals appear to take precedence over social needs, and where access to piped water is conditional on the increased policing of certain individuals and spaces.

V. The Invisibility of Urban Water Networks

This paper has examined the increased visibility of urban water networks in both poor and non-poor areas, and has argued that the concessionaires engage in such practices in part to reduce NRW. In low-income areas, the techniques directed at curbing non-payment predicate access to water upon the increased policing of certain citizens and spaces, and that policing has partially been transferred from the concessionaires to individuals, CBOs and entrepreneurs, thus shifting the politics of non-payment closer to the community level.

But while these processes increase visibility for the concessionaires and affected communities, a parallel process is at work – one that masks the imperfections of current modes of infrastructure expansion. As the concessionaires produce and propagate aggregate statistics in the national and international arena, households that are unserved or underserved remain obscured in two ways. First, the aggregate data do not distinguish between types of connections – direct, clustered or bulk – even though there are financial and behavioural implications associated with each. Second, the concessionaires do not define coverage consistently – at times, it refers not to actual connectivity but to the possibility of connection. Thus, it is most likely that those who do not connect to the utilities – because connection fees are prohibitive or for other reasons – but who live close to a main line are considered covered. In Santa Ana, for instance, Maynilad records reflect that the two bulk meters serve 800 households because that is the approximate number of households in the vicinity.⁽⁴⁴⁾ In reality, the micro network currently serves about 300 households, with the remainder using alternative sources. A similar overestimation became apparent during the initial post-privatization years, when both concessionaires were using a multiplier of 9.2 persons per household.⁽⁴⁵⁾ Because the same multiplier was used for all cities in Metro Manila, it was later discovered that there were some cities where the coverage claimed by the concessionaires exceeded official census numbers.

The Satterthwaite critique of some plans to meet the Millennium Development Goals cautions that an overreliance on aggregate statistics can be problematic because it tends to underestimate the scale and particularities of urban poverty and inequality.⁽⁴⁶⁾ More generally, Rose finds that numbers are politicized judgments that reflect what and how to measure and the ways in which we interpret those results.⁽⁴⁷⁾ In large metropolitan areas, the use of aggregate statistics simultaneously helps to clarify and obscure the complexity of urbanization. Arguably, the inflation of statistics can be tied to financial motivations, political success or other perverse incentives that detract from actual poverty and inequality alleviation. This is not to say that metrics such as coverage and NRW do not contribute to our understanding of urban processes; rather, equal attention should be paid to the details that the numbers omit, as to what they purportedly represent.

In Manila’s case, the dual processes of rendering networks visible and invisible have four implications:

VI. Re-Thinking Incentives

The challenges of urban water provision, particularly to low-income households, have been demonstrated in multiple ways. Budds and McGranahan point to the incompatibility of large public and private providers in serving most low-income households.⁽⁴⁹⁾ Relatedly, Wolff and Hallstein suggest that neither public nor private utilities have an inherent efficiency advantage and that decisions need to be made on a case-by-case basis.⁽⁵⁰⁾ Bakker et al., as well as Loftus and McDonald, examine the difficulties of water management when there is regulatory capture, while Allen et al. make the case for the need to re-visit incentives for governance and water provision.⁽⁵¹⁾ Recognizing the need for alternative incentives, Hardoy and Schusterman suggest applying a higher weighting coefficient to the service of low-income areas.⁽⁵²⁾

Manila’s concessionaires have overcome some of the challenges of pro-poor provision, particularly in Manila Water’s jurisdiction. However, current approaches continue to be problematic in their overestimation of increased coverage and downplaying of remaining inequities. The differentiated terms for accessing water for some households via micro networks or clustered metering, as well as for those that remain completely unconnected, serve as boundary cases that illustrate the limitations of current modes of infrastructure expansion. If universal water access is our end goal, then we must find ways for the public to access the concessionaires’ rich, on-the-ground knowledge, so that we can better examine the barriers that prevent the unserved and underserved from obtaining water on the same, or better, terms as the non-poor. Manila Water’s “walk the line” strategy, for example, provides managers with detailed house-to-house information, yet most of the disaggregated data for households that are not directly connected – that pertaining to indirect users, non-consumers or disconnected ones – do not get passed on to policy makers. A long-term shift in water management might involve a complex shift in incentives, such that the utilities and their private agents would be pressured not only into becoming “model utilities” but also into producing accurate data to be shared and analyzed. More realistically, however, there are some direct ways of addressing remaining barriers in Manila. While strengthening the MWSS Regulatory Office might be one way to improve monitoring efforts, the Philippines’ rich civil society sector could also serve as independent reviewers. For instance, the passage of a Freedom of Information Bill might allow interested civil society parties to review reports that the concessionaires submit to the regulatory office.⁽⁵³⁾ Alternatively, third-party audits of the concessionaires’ progress could include surveys of households that are not connected, as opposed to the current system of evaluation in which only a small percentage of serviced households are asked for feedback.⁽⁵⁴⁾

In cities such as Manila, access to water is often tied to broader structural concerns such as competing claims for land and a general lack of economic opportunity. Furthermore, rural and smaller urban areas tend to receive less political and financial attention, which can translate into reduced access to basic needs. Focusing on these realities sometimes leads to an oversight of the remaining unserved and underserved in primary cities. In order to determine where limited resources must be spent, however, it would behoove policy makers to obtain more precise information on barriers to access – information that is, at times, already available, yet not dispersed. The result of this effort would hold the evidence-based promise of rendering the underserved visible and subsequently available for inclusion in attempts to universalize access to water.