Pro-poor bottled water? : the role of refill water in providing inclusive access to water for low-income communities in Jakarta, Indonesia

(1)

Pro-poor bottled water?

The role of refill water in providing inclusive

access to water for low-income communities in

Jakarta, Indonesia

Carolin Tina Walter

10256040

carolintinawater@gmail.com

Thesis in partial fulfilment of the requirements of the Research Master

International Development Studies (MSc), Department of Geography, Planning,

and International Development, Graduate School of Social Sciences (GSSS),

University of Amsterdam

Date of Submission: 1

st

_{of August, 2016}

Supervisor: Dr. Michelle Kooy, UNESCO-IHE Institute of Water Education, Delft

Local Supervisor: Indrawan Prabaharyaka, Technische Universität München

Second Reader: Dr. Michaela Hordijk, University of Amsterdam

(2)

i

Acknowledgements

It is with outmost sincerity that I express my gratitude to the following people, who supported me throughout the past year and more in bringing this thesis to reality. First and foremost, I am deeply thankful to my supervisor, Dr. Michelle Kooy, who guided me through this journey from beginning to end and supported me in any way a scholar-to-be could hope for. Without her engagement, kindness, and brilliance this study would not have been the same.

This research would not have been possible without the help of my local supervisor, Indrawan Prabaharyaka, who has guided me in the field and after with competence and friendship, both in my academic endeavours and in finding my (figurative and literal) way in Jakarta.

I am deeply grateful to Melinda Martinus and Vincent Pooroe, my translators, advisors, mentors, and friends, who have supported this research with great enthusiasm throughout the entire time in the field. Likewise, I extend my sincere gratitude to Budi Darmawan for his keen interest and engagement with my project.

I thank all interview respondents and survey participants for their contribution to this research.

I thank Emma and Roby, as well as Khansa and the rest of the Taman Suropati family, for turning Jakarta from a site of fieldwork into a place I called home with conviction. With all my heart I thank Tami and Samuel, whose selflessness and kindness I will never forget.

This list would not be complete without thanking the Social Crew for being by my side every step of this journey.

Finally, I am grateful to my family for their love, encouragement, and constant faith in my capabilities.

(3)

ii

List of Abbreviations

ADB Asian Development Bank

Aetra PT Aetra Air Jakarta

Bappenas Indonesian Ministry of National Development Planning (Badan Perencanaan Pembangunan Nasional)

Bppspam Agency for the Support of the Development of the Drinking Water Distribution System (Badan Pendukung Pengembangan Sistem Penyediaan Air Minum)

BPS Central Statistics Agency (Badan Pusat Statistik) IUWASH Indonesia Urban Water, Sanitation, and Hygiene

JMP WHO & UNICEF Joint Monitoring Programme for Water and Sanitation

MDG Millennium Development Goals

OECD Organization for Economic Co-operation and Development

Palyja PT PAM Lyonnaise Jaya

PDAM Local water company (Perusahaan Daerah Air Minum) POU Point-of-use Treatments

Puskesmas Community Health Centre (Pusat Kesehatan Masyarakat)

RO Reverse Osmosis

RPJMN National Medium-term Development Plan 2015-2019 (Rencana Pembangunan Jangka Menengah Nasional)

RW Refill Water

SDG Sustainable Development Goals

SSIWP Small-scale Independent Water Provider

STMB National Community-based Sanitation Programme (Sanitasi Total Berbasis Masyarakat)

USAID United States Agency for International Development WASH Water, Sanitation and Hygiene

WB World Bank

WSP Water and Sanitation Programme (World Bank) WSS Water and Sanitation Sector

(7)

vi

List of Tables and Figures

Table 3.1 | Demographic overview of the interviewed refill-water vendors Table 3.2 | Overview over numbers and types of interviews.

Table 4.1 | Connected and non-connected refill water users Table 4.2 | Sources of drinking water in Jakarta

Table 4.3 | Reasons for purchasing refill water

Table 4.4 | Monthly household income per research location and total Table 4.5 | Correlation of poverty indicators

Table 4.6 | Purposes of use of refill water

Table 4.7 | Household’s combination of multiple water sources

Table 5.1 | Customer perceptions of quality of different water sources Table 5.2 | Average price per unit water sources

Table 5.3 | Customer perceptions price of different water sources

Table 6.1 | Monthly refill water expenditure and consumption per income level Fig. 3.1 | Conceptual scheme

Fig. 3.2 | Administrative categorisation of the research locations.

Fig. 3.3 | Research locations in the northern (Penjaringan) and Southern (Gedong, Tengah) parts of Jakarta

Fig. 3.4 | Data collection timeline

Fig. 3.5 | Cycle of data collection and analysis as used throughout fieldwork. Fig. 4.1 | Service coverage Palyja and Aetra combined, 1998-2013

Fig. 4.2 | A refill water station in Tengah Fig. 4.3 | Patterns of water sources accessed

Fig. 5.1 | Scheme of a refill water processing installation through reverse osmosis Fig. 5.2 | Customer perceptions quality refill water

Fig. 5.3 | Customer perceptions price refill water

Fig. 6.1 | Refill water transportation trucks being filled with raw water in Cibinong, Bogor.

(8)

vii

Abstract

Bottled water consumption is growing rapidly across cities of the Global South. In International Development, the equation of water access with access to ‘improved’ water sources has predominantly led to analyses of equity of access to the centralised water network, rather than within or beyond it. Informed by perspectives of Southern Urbanism, this research expands the study of equity of water access to bottled water, by investigating the role of refill water (air isi ulang), a particular form of non-branded bottled water, in contributing to inclusive water access for low-income households in Jakarta, Indonesia. Employing an Inclusive-Development lens and focusing specifically on low-income communities, this study contributes to the discussion concerning pro-poor dimensions of urban water provision. Over a period of four months, empirical data was gathered in Jakarta, following an exploratory sequential mixed-methods research (MMR) design. Quantitative data is gathered from a household survey in two low-income neighbourhoods in Jakarta, and qualitative data in form of in-depth interviews with refill water vendors, and with key informants in water and international development sectors. This study finds that refill water fills a gap left by fragmented, intermittent, low-quality piped water supply, non-potability of alternative water sources, and high price of branded bottled water. It also identifies how the adherence to an ideal of universal piped-water coverage by government and the sphere of international development in Jakarta, has led to a paradoxical situation, where refill water is growing in reach and importance for lower-income communities, all the while being left largely unregulated, creating uncertainties surrounding water quality. This study finds that lower-income refill water customers perceive water as best value-for money option available to them, thereby emphasising the need for a differentiation between drinking water sources and water used for other purposes. Overall, the example of refill water highlights the need for a more differentiated conceptualisation of water access for pro-poor approaches, which understands as normal the heterogeneous configurations of urban water in cities of the Global South, rather than a pathology of an ideal, centralised network-to-be. This study argues that expanding our understanding of equity of water access beyond the piped network is necessary when aiming to create inclusive cities in the context of the Sustainable Development Goals (SDGs) and the current and future international development agenda.

Keywords bottled water, SDGs, Inclusive Development, Southern Urbanism, MMR, Jakarta

(9)

1

Introduction

Approximately 663 million people lack access to safe and clean drinking water (United Nations, 2016). This is despite the fact that globally, the target for drinking water to ‘halve, by 2015, the proportion of the population without sustainable access to safe drinking water and basic sanitation’ was achieved five years ahead of time in 2010, and 2.6 billion people have gained access to water since 1990 (ibid.). The global achievement of this target however conceals considerable differences both between and within countries. What is more, the achievement of this target is in doubt – meaning fewer people gained access than reported – because of problems in the indicators: firstly, they are based on the definition of access to ‘improved water sources’, which describes types of sources1_{, rather than the quality of the water (Onda, LoBuglio, & Bartram, 2012) or the} quality of this access (Burt & Ray 2014; McIntosh, 2014); and secondly, the existence of these improved sources does not reveal anything about who is able to access these sources, under which circumstances, or if people actually make use of them at all. As a result, in many places where this target has been achieved, water is still not reliable or safe.

While ‘improved’ sources might not deliver safe water access, one ‘unimproved’ source might: initial research on various types of bottled water2_{finds that they play a role in} providing improved access to drinking water in cities across the developing world. In Ghana, packaged ‘sachet’ water increased substantially the proportion of people able to access (reasonably) safe water and gaining the associated health benefits (Stoler, 2012), and in Indonesia, use of non-branded bottled water, locally called refill water (air isi ulang), is associated with the reduction of childhood diarrhoea in comparison to the available tap water (Sima et al., 2012). In light of this mismatch between indicators and realities of access, there is a need for a more nuanced assessment of water access – and indeed, the post-2015 development agenda in the form of the Sustainable Development Goals (SDGs) has made some adjustments vis-à-vis its predecessors. Now, one out of the seventeen Goals is entirely devoted to clean water and sanitation (Goal 6), and its sub-division into eight Targets addresses the complexity of water access beyond the existence of ‘improved’ water sources. It includes attention to the quality of water accessed, and the way this quality is influenced by larger hydro-social interactions, like pollution and wastewater contamination. Likewise – and importantly – Goal 6 also includes a focus on equity of access, a dimension of water access, which was entirely brushed over by the MDGs’ indicators.

1_{Improved sources, according to the WHO/UNICEF Joint Monitoring Programme, are ‘piped}

water, public tap, borehole or pump, protected well, protected spring or rainwater’ (WHO/UNICEF, 2015).

2_{Throughout this paper, bottled water means all forms of packaged, purified drinking water, both}

(10)

2 The achievement of the MDGs and the relevance of the SDGs are most problematic in cities of the Global South. It is predicted that 66% of all of the world’s population is going to live in urban areas by 2050 (World Bank, 2015), yet access to water infrastructure in cities remains a major challenge: at the moment, population growth in urban areas is already outpacing current infrastructure needs and as a result, overall access to urban piped water networks has stagnated between 1990 and 2015 (WHO/UNICEF, 2015). At the same time, rapid urbanisation in combination with climate change is one of the key drivers for increasing water demand in the near future: within the next three decades, ‘the demand for water in cities is projected to increase by 50 to 70 percent’ (World Bank, 2015, p. 27). Solutions aiming to expand urban water access have largely remained the same over time, involving what has been called the ‘infrastructural ideal’: the distribution of treated water through a large-scale centralised piped network, with the ultimate goal to provide potable tap water to all urban residents (Graham & Marvin, 2001).

However, this goal is out of reach of many countries and might never be achieved. Some areas have even witnessed a decrease in water access since 1990, such as Sub-Saharan Africa, where access to piped water has decreased by 10% (World Bank, 2015). In addition to this infrastructure deficit, where networks do exist, they are often unreliable and only provide intermittent water. Likewise, the quality of piped water throughout the network is variable and influenced by an assemblage of multiple factors, ranging from environmental dimensions, such as impact of the seasons and raw water availability and quality, to technical and practical issues, such as old pipes and pressure drops. Thus, as the world continues to heat up, develop economically, and urbanise, current water access and supply patterns will be unable to meet growing cities’ water needs in the near future, leading to serious concerns over urban water security, especially for the urban poor (ibid.).

Deficient infrastructure, delivering poor-quality water, is leaving large gaps in many urban water networks in cities of the Global South. A growing number of scholarly work indicates that this gap is increasingly filled by various types of bottled water, involving branded bottled water, but also other forms of bottled water, such as packaged ‘sachet’ water in West Africa (Stoler et al., 2012; Stoler, 2015) or refill water in Indonesia (Sima et al., 2012) and the Philippines (Francisco, 2014). Bottled water is increasing in importance for households of all socioeconomic classes and across regions – nevertheless, there are indications that there might be different motivations for consuming bottled water for different (income) segments of society. While poor tap-water quality appeared to be the main reason for bottled-water consumption for people with higher income, poorer people indicate they consume bottled water because of lack of alternatives (Sharma & Bhaduri, 2014, p. 6). In Ghana, packaged water in the form of sachets ‘may be one piece of a decentralized drinking water provision strategy for low-income urban communities’ (Stoler et al., 2015, p. 60). In Indonesia, initial assumptions assumed that while higher-income households consume branded bottled water, poorer households make use of the non-branded, cheaper, refill water. However, it is now clear that refill water is consumed by people of all socioeconomic classes (World Bank, 2015a).

(11)

3 Within cities, bottled water is sold by a multitude of brands and non-branded versions of varying price and quality (Sharma & Bhaduri, 2014; Ahmed et al., 2013) and its production has considerable ecological and social impact beyond urban areas, for instance, by impacting groundwater levels (AMRTA, 2014; KRUHA, 2012).

However, this growth of the bottled water sector has largely been ignored in the context of international development and urban water provision. Where it is acknowledged, it is seen as a ‘failure’ of the development project (WSP, 2015). There seems to be the assumption – like for vended water or so-called informal providers – that it will phase out over time when (if) the urban infrastructural ideal is achieved (World Bank, 2015). For instance, all types of bottled or packaged water count as ‘unimproved’ water sources and the common assumption is that bottled-water-consumption arises from a lack of ‘improved’ alternatives – piped water most importantly. However, what research there is on bottled water suggests that consumption of bottled water does not necessarily follow from a lack of access to the formal network (Francisco, 2014; Stoler et al., 2015). As a result, bottled-water consumption may not fade away, even in areas where access to central water infrastructure is – or will be – provided.

The growth of the bottled-water sector is especially rapid in Indonesia: between 2009 and 2014, the country’s bottled-water sector grew at a pace of more than 12% per year and 23.1 billion litres of bottled water were consumed nationwide in 2014 (Prasetiawan, 2015). This rapid increase makes Indonesia the second largest market for bottled water in Asia after China (IBWO, 2014; Warburton, 2011) and counts more than 400 brands of bottled water on its markets (Hadipuro, 2014). Within Indonesia, the capital city of Jakarta, whose extended metropolitan area totals more than 28 million inhabitants, represents a majority of this growth in bottled water provision. Here, the share of people relying on some form of bottled water (branded and/or refill) amounts to at least 71% currently, and is predicted to keep growing steadily (BPS, 2015; Rani et al., 2012). Presumably, such rapid expansion profoundly impacts and transforms the landscape of urban water provision by providing water, which is of potentially higher quality than alternatives, yet considerably more expensive per unit volume than for instance piped water or groundwater. Especially the cheaper, non-branded refill water might widen the socioeconomic range of people accessing bottled water. However, bottled water has received little to no attention within the field of international development scholarship and practice. If at all, it is defined, as other sources of water, through what it is not – refill water is ‘non-piped’, ‘unimproved’, and an ‘alternative’. As such, this classification reflects the long-upheld ‘infrastructural ideal’ as gold standard of water provision in cities of the Global South (Graham & Marvin, 2001). However, the idea that this model of water provision can be replicated in urban areas of middle-and-low income countries has been questioned repeatedly, and is most explicitly challenged by the fact that most cities in these countries have not yet reached universal coverage, and have made very little progress in this regard throughout the past decade(s) (Jaglin, 2014; Parnell, 2014).

(12)

4 Assumptions as to the transitional nature of the bottled water market, and temporary reliance on bottled water for drinking water sources by Indonesian residents, also informs Jakarta’s approach to urban water provision. From this perspective, refill water and bottled water are used as an argument exemplifying the disadvantages those not connected to centralised network experience as a consequence of their exclusion, such as a much higher price paid for their water (e.g. McIntosh, 2014). While these are valid concerns, examining refill water simply as ‘non-piped’ source is not sufficient to understand its precise role in the landscape of urban water provision, when it is accessed and why, whom it is providing access to and under what circumstances, or the ways in which it can be inclusive. Is it a beverage consumed by choice; is it an indicator of inequality in access to piped water; or is it a rational response to the options available to low-income residents? These are some of the areas this thesis research is going to explore. Therefore, this paper investigates refill water in the context of Jakarta by asking the following question: What is the role of refill water in providing inclusive access to water supply in Jakarta, Indonesia?

This is especially crucial in this context of the adoption of a new set of global development goals. While the SDGs do offer a more differentiated approach to water access than the MDGs did before them, bottled water is still conceptualised as merely another ‘unimproved’ source, leaving the international development community at risk of ignoring once again a form of water provision, which is likely to considerably impact the landscapes of urban water provision and equity of access in cities of the Global South.

The purpose of this paper is thus an analysis of the refill water sector in Jakarta by means of an exploratory Mixed-Methods Research (MMR) design through the theoretical lens of Inclusive Development and Southern Urbanism. Thereby, it hopes to gain an understanding of the role refill water plays for water access in Jakarta, identify why it is not yet sufficiently recognized within international development and studies on the urban water sector, and explore how in the context of Jakarta it is relevant for the achievement of goals of Inclusive Development, and in the context of the SDGs.

This paper is a Master’s thesis, following research conducted as part of the Research Master Programme in International Development Studies at the University of Amsterdam. It is structured as follows: first, the theoretical framework is established and the research contextualised in its specific location; then, methodology and data-analysis processes are described, before proceeding to present results in three thematic sub-chapters. Next, these results are discussed in light of debates on Inclusive Development and Southern Urbanism, and some recommendations are made on bottled water in the SDG-era.

(13)

5

Theoretical Framework

This chapter provides the theoretical framework for this study and a review of relevant literature on the topic of bottled water. First, this chapter provides a review of the way bottled water has been discussed in academic literature so far, with specific focus on literature within international development. Then, current debates concerning urban water access from a perspective of Southern Urbanism are discussed. Finally, Inclusive Development is introduced as a framework for the pro-poor focus of this study.

2.1 Bottled Water

Bottled water consumption is growing globally – it has evolved from a niche market providing ‘mineral’ water to a selected number of high-end consumers to one of the fastest growing industries in the world (Greene, 2014). While global expansion of the bottled-water market was especially quick before the year 2000, it has been growing at a steady pace of 6% annually ever since and is expected to keep growing in the future (Rani et al., 2012; Younos, 2013). In 2014, global consumption of bottled water was more than 282 million litres (IBWO, 2014). This trend is observable in countries all across the globe, rendering bottled water the second most consumed beverage worldwide (Rodwan, 2011).

Despite this ubiquity, the topic of bottled water is covered surprisingly scarcely in academic literature, especially in the social sciences. Most of existing literature on the topic tries to understand the paradox that is bottling water – how come people started to pay for something that, essentially, falls from the sky for free? How is it possible that ‘a billion-euro industry has been created around selling a product with little or no distinguishing features? ´ (Collins & Wright, 2014, p. 27). These studies usually have a background in disciplines like business and consumer studies, and investigate the ways in which marketing strategies have contributed to this expansion (e.g. Collins & Wright, 2014; Massoud et al., 2013; Viscusi, Huber, & Bell, 2015; Wilk, 2006).

In these studies, the authors agree on three main drivers for consumers to opt for bottled water: perceptions of convenience, taste, and safety of the water. Most of these studies are based on a comparison of bottled water with tap water (e.g. Viscusi, Huber, & Bell, 2015). Doria (2006), among others, found that consumers in the U.S, Canada, and France are dissatisfied with the taste of tap water and therefore opt for the bottled version. This is a phenomenon predominantly research conducted in middle- and higher-income countries indicates. When respondents in lower-income countries indicate ‘taste’ as a reason for choosing bottled over tap water, it is often as a proxy for the assessment of the water quality, and therefore usually correlates with the indication of ‘safety’ as major decisive factor (Francisco, 2014; Massoud et al., 2013). Mistrust in tap water is one of the most important drivers of bottled water consumption, and the perception that bottled

(14)

6 water is of superior quality than tap water is pervasive cross-culturally (Greene, 2014; Jaffee & Newman, 2012). However, studies evaluating this claim are of mixed results, indicating varying degrees of quality and safety of bottled water across and within regions, thereby challenging the claim of superior quality of bottled water. While bottled water might not only not be of better quality than tap water from Italy (Cidu, Frau, & Tore; 2014) to the Middle East (Massoud et al., 2013), in some cases it might actually be worse: in Greece, for instance, close to 14% of tested bottled water did not comply with the national water regulations (Venieri et al., 2006). Ghana’s sachet water quality is variable (Oyelude & Ahenkorah, 2012), as is bottled water in Dhaka (Ahmed et al., 2013), and some studies raise concern about the PET bottles themselves and the travel of contaminants from the plastic into the water in Europe and Africa (Tukur et al., 2012). This raises concern about another issue: A number of studies raise concern about the lack of regulation for bottled water, which tends to be much stricter for municipal water utilities in lower-income (Hadipuro, 2010; Massoud et al., 2013) and higher-income countries (Hu et al., 2011; Jaffee & Newman, 2012). This deregulation is also problematic in in terms of environmental sustainability: bottled water production (e.g. Gleick & Cooley, 2009) and the plastic waste produced (e.g. Saylor et al., 2011) go largely unregulated, leading to increasing concern about environmental sustainability.

Some authors have taken their analysis a step further and challenged the dominant narrative of bottled-water consumption based on universal perceptions of superior quality, backed by a successful marketing machinery. Sharma & Bhaduri (2014) point towards the importance of specific cultural contexts, which were different ones, for instance, for the emergence of bottled water in Northern America than in Europe. With specific reference to lower-income countries, the authors suggest that ‘this line of argument undermines the importance of extra-market forces and regulatory structures, such as personal experiences, influence of peer groups, trust on science-based technologies, regulatory governance, and trust on various institutions (public/private) involved in supplying drinking water’ (Sharma & Bhaduri, 2014, p.3).

And indeed, since bottled water emerged first in the Global North, most research focuses on those regions. However, these patterns are shifting – while bottled water consumption has been around the longest in Europe and North America, the market has been expanding rapidly in middle- and low-income countries since the early 1990s. Today, the five largest markets for bottled water are, in order, China, the United States, Mexico, Indonesia, and Brazil (IBWO, 2014). Following traditions of research, which had already guided studies in Northern countries before, researchers, who did investigate reasons for this expansion, did so by expanding these theories to these new markets in lower-income countries. This led some to conclude that, for instance, this expansion is due to a growing middle class in these countries (Rani et al., 2012). However, as explored in more detail in the following sub-section (2.2), a growing body of scholarly research points to the unique context of cities of the Global South, asking for new analyses and theories on urban water access grounded particularly in Southern urban experiences (Jaglin, 2014).

(15)

7 By far the most dominant argument is that in cities of the Global South, bottled water consumption is a response to inadequate urban water infrastructure (Wilk, 2006). Taking specific reference to the piped network, a number of studies conclude that ‘failure of the established public sector provisioning mechanisms compels certain sections of people to consume bottled water’ (Sharma & Bhaduri, 2014, p.6). In Mexico, Greene (2014) suggests that ‘general insecurity in the public water system’ has fostered bottled-water consumption (p. 1). Some frame this as an opportunity: Younos (2013) suggests that bottled water can, under circumstances, provide a form of decentralised water provision in low in-come area as a (partial) solution to the global lack of safe drinking water crisis (Younos, 2013, p. 224). However, as Greene (2014) argues, ‘the provision of bottled water as a solution to the water crisis may come at the expense of the reinvestment in the public water infrastructure’ (pp. 15-16). Not only might it provide disincentive for governments to expand their piped network, it might also curb their capability of doing so: for instance, when richer households stop using piped water, they prevent cross-subsidisation to the urban poor (Hadipuro, 2010).

This argument falls into wider debates surrounding the topic of ‘water privatisation’ in cities of the Global South, which has been subject to fierce academic debates since it has been pushed as major water-provision strategy in developing countries since the early 1990s (see section 2.2.1). While these debates almost exclusively surround private-sector participation in urban water utilities, Jaffee & Newman (2013; 2013a) view the growth of bottled water as another expression of the privatisation of water and commodification of nature. According to the authors, bottled water does not encounter the same obstacles to capital accumulation that municipal utilities have, when subjected to ‘water privatisation’– rendering it a ‘more perfect commodity’ (Swyngedouw, 2005). The authors conclude that this nature of bottled water renders it ‘a more serious threat to the project of universal public drinking water provision than that posed by tap water privatization’ (Jaffee & Newman, 2013, p. 1). Thus, while authors remain in disagreement whether it should, they do agree that bottled water does fill a gap in urban water provision left by inadequate infrastructure.

This is problematized by international development institutions and studies on poverty and urban inequality. In India, ‘people citing the unavailability of tap water as the most important reason for bottled water consumption are primarily coming from lower family income groups’ (Sharma & Bhaduri, 2014, p. 6). In light of the fragmented infrastructures of many Southern cities reaching predominantly well-off households, this implies a much higher proportion of income paid for water by poor households (Bakker et al., 2008, p. 1896). However, at the same time, studies have challenged this straightforward relationship between lack of access to the water network and bottled water consumption and called for more differentiated analyses (Bakker et al., 2008; Kooy, 2014). Studies from the Philippines (Francisco, 2014), Jordan and Lebanon (Massoud et al., 2013) and Ghana (Stoler et al., 2015) show that even in cities with (near-)universal coverage, bottled water consumption is important and growing. Thus, while there is an interaction between bottled-water consumption and quality and

(16)

8 consistency of networked water services water across countries, there are indications for regional differences. Dissatisfaction with tap water stems predominantly from taste concerns in many high-income countries, such as the Netherlands and Australia (Johnstone & Serret; 2012), while in Mexico, the Philippines, Lebanon, and Jordan, these concerns revolve around health and water safety (Francisco, 2014; Greene, 2014; Massoud et al., 2013). Thus, reducing the argument to access to piped water might be misleading and insufficient. Rather, as Sharma and Bhaduri (2014) explain at the example of Delhi, India, the poor quality and unreliability of piped water might be a major factor influencing bottled water consumption.

In addition to piped water access, the role of income too is as of yet inconclusive. In Metro Cebu, the Philippines, a study of 360 households found no indication that income plays a role for the likelihood of consumers to buy bottled water (Francisco, 2014). An OECD study finds contrasting results: Negative perceptions of tap water quality influences the decision to purchase bottled water to a much larger extent than the decision to resort to water purification systems in OECD countries, where income does have a small positive impact on the adoption of bottled water consumption (Johnstone & Serret, 2012). This is another indication that motivations for bottled water consumption might differ between higher-income and lower-income countries3_{. These examples} indicate that expanding analyses based on the experience of Northern cities onto the South might result in at best incomplete conclusions, and resonate with the call for new, differentiated approaches to dynamics of water access and provision based on evidence from the South.

The focus of this study is Jakarta, Indonesia. Its largest bottled-water brand, Aqua by Danone, entered the country in 1974, but market expansion gained particular traction after the turn of the century (Hadipuro, 2010). Here, consumption of bottled water has been growing at a steady rate of more than 12% between 2009 and 2014 (IBWO, 2014; Prasetiawan, 2014). In 2014, more than 20 billion litres of bottled water were consumed nationwide in Indonesia, making it the world’s fourth largest market for bottled water (IBWO, 2014).

This thesis focuses specifically on a particular type of non-branded bottled water, locally called refill water (air isi ulang), which is sold via decentralised small-scale refill stations and individual vendors, selling non-branded, purified, non-networked water from outside the city. Refill water serves ever-increasing proportion of urban populations across Indonesian cities. In Jakarta, for instance, branded bottled water accounts for about 48% of total bottled-water consumption, while refill water serves the other 52% (BPS, 2014). The phenomenon of refill water is not exclusive to Indonesia, and there are instances of refill water types in the Philippines (Francisco, 2014) and Thailand (Hadipuro, 2010). However, while this type of packaged water serves an increasingly large share of populations in urban areas of the Global South, it has attracted barely any scholarly attention within studies on the urban water supply sector and access to water

3_{A study by Hu, Morton, & Mahler (2011) found that perceptions of low safety of municipal water}

(17)

9 services. Those studies mentioning refill water do so raising the same concerns of high price-per-unit and disproportional financial burden carried by the urban poor as a result of a lack of access to piped water, just as literature on (branded) bottled water does. While initially it was assumed that refill water would mainly attract lower-income customers, research shows that refill water is consumed by people of all socioeconomic classes (World Bank, 2015a, see Fig. 1.1). Sold via small-scale private vendors, refill water is located conceptually at the intersection between literature on small-scale water vendors (‘purchasing water from local vendors’) – but these usually function as an extension of the piped network – and bottled-water literature – which usually focuses on branded bottled. Thereby, refill water does not match preconceived conceptualisations of bottled water or ready-made theories to evaluate its role.

Fig. 1.1| Bottled water usage by income quintiles in urban areas, 2009-2013. Source: World Bank (estimates; 2015).

As a result, bottled water in general and refill water in particular are poorly understood in academic literature. Particularly studies in the field of international development and urban water access have largely failed to address this rapid expansion of bottled water in urban areas of cities of the Global South. It has mainly been utilised as an example of the consequences of inadequate urban water infrastructure and brushed off as ‘unimproved’, thereby foregoing chances to assess positive and negative potentials of refill water on inequalities in access. No study on urban areas has evaluated how bottled water access is produced by an urbanising population and how it shapes the landscape of urban water. At the same time studies have yet to address the way in which bottled water interacts with other sources of (drinking) water in a heterogeneous landscape of urban water provision in cities of the Global South. This ovrsight repeats the lack of attention to other non-networked sources (for instance groundwater), and informal providers (for instance water vendors), who were also viewed as ‘transitional’ despite their persistent ubiquity (Ngangyuka et al., 2014).

The purpose of this study is thus twofold: one, it aims to explore a type of packaged water not much is known about other than minor attention from market/business analysis and public health/water quality analyses; and two, by doing so, it emphasises the need for Southern Urbanism and lenses appropriate for cities of the Global South, particularly with reference to a shift in global international development in form of the MDG-to-SDG transition, which is addressed in the following section.

(18)

10

2.2 ‘Access to’ Water in Cities of the Global South

‘Public’ or ‘Private’ or Both or None?

Throughout at least the past two decades, a large majority of the debate surrounding access to urban water in international-development scholarship and practice has surrounded the relative merits and drawbacks of the ‘public’ or ‘private’ sector in expanding the piped-water network in cities of the Global South, particularly those analyses concerned with equity of access. Although an increasing number of studies now question the merits of that debate, as elaborated below (see section 2.2.2), it has shaped much of the framework of research and development practice in Southern cities in the recent past, which is why it is briefly summarised in this sub-section.

At the point of independence, the young governments of many former colonies were struggling to cope with the colonial legacy of fragmented, often inadequate water supply systems in their urban areas (Bhatt, 2014; Kooy & Bakker, 2008). Some authors suggest that ‘[i]n many countries the institutional capacity to plan, implement and administer large-scale urban infrastructure vanished along with the exodus of the colonial leadership’ (Bhatt, 2014, p. 103). Exacerbated by the alleged lack of experience and efficiency of developing countries’ public sectors, political instability, and other problems, such as corruption, little to no progress was made in the expansion of piped-network coverage (Castro, 2008), leading critics in the early 1990s to argue that the public sector had failed to live up to its promise of providing universal coverage, and that alternatives, specifically in the form of private sector participation (PSP), had to be considered (e.g. Winpenny, 1994).

What ensued is one of the most fiercely debated approaches to water provision in cities of middle-and-low-income countries, gathered under the concept of ‘water privatisation’. As disappointment in the public sector was paired with optimism concerning market mechanisms for effectively allocating resources, the 1990s witnessed increasing mainstreaming of policies promoting private actor involvement for the solution of water and sanitation problems in developing countries (Bakker, 2003; Castro, 2008). Vigorously promoted by international development institutions, most notably the World Bank (Goldman, 2007), this went hand in hand with a decrease of public ownership of and control over water utilities and a conceptualisation of water as an economic good rather than as primarily a public good. Water should be allocated through the market instead of the public sector, it should be provided by private operators, while its regulation by governments should be minimised or abolished (Castro, 2008).

Arguments which were especially powerful for both sides of the debate were centred around the urban poor: the advocates of privatisation considered it a pro-poor solution (e.g. Winpenny, 1994), while opponents suggested that private sector involvement negatively affects precisely poorer households, because they might not be able to access water anymore once they have to pay for it (e.g. Swyngedouw, 2005). However, the public-private binary has increasingly been contested as constructed (Bakker, 2013), as a debate, which is ‘missing the point’ (Budds & McGranahan, 2003), and insufficient to

(19)

11 explain the continued lack of access to water in cities of the South (Bakker, 2007; Bakker et al., 2008). Budds and McGranahan (2003) call attention to barriers beyond economic ones inhibiting poor households to connect to the water network, such as precarious land tenure in informal settlements. The authors suggest that ‘many of the problems that have been encountered with privatization can also arise with public utilities, while many of the strengths of private sector participation can also be achieved by reforming public sector utilities’ (p. 112). In Jakarta, for instance, faulty institutions might play an important role in determining the continued fragmentation of the water network, by creating disincentives (of both public and private actors) to expand connections to poorer neighbourhoods, and of the poor themselves to strive to connect to piped water networks (Bakker et al., 2008).

Budds and McGranahan (2003) make another important point, namely that ‘the scale of attention to privatization in recent years somewhat obscures the fact that the majority of the population in the South continues to be served by the public sector or small-scale or informal providers’ (p. 111). As pointed out in the previous section (2.1), refill water too is sold via small-scale private vendors, making them part of what loosely falls within ‘private sector water provision’. Nevertheless, they are largely left out of this debate. These arguments highlight the need to move beyond over-simplified dichotomies and narrow definitions of private-sector involvement in water provision. The following section does so by outlining most prominent debates concerning water access in cities of the Global South. It expands the argument made in the context of bottled water, namely that analyses of urban water in the South need to move beyond infrastructure-centred approaches and other idea(l)s grounded in experiences of Northern cities.

Beyond the Dichotomy: Southern Urbanism and Pro-Poor Water Access in Cities of the Global South

As abovementioned dichotomous theories remain at a loss of explaining comprehensively the persistent disconnection of urban residents in general, and the urban poor in particular from water access in Southern cities, the voices of critical scholarship are gaining increasing momentum. Their common point: Maybe we are asking the wrong questions. In academic literature as well as in the approaches of the large development institutions, ‘access’ to water in urban areas has almost exclusively been conceptualised as access to the formal water network (e.g. Bakker, 2003; Tan, 2012). Therefore, most studies on ‘pro-poor’ urban water access strategies investigate the reasons and effects of the poor being excluded from piped water access. Analyses of equity have taken the piped network as point of reference and departure, which is why proposals for increasing equity of access are usually – and logically – based on adjusting the functioning or governance of the centralised water infrastructure. This means that analyses of equity of water access in the Global South within the piped network (e.g. different modalities of access, such as direct or indirect connections), and outside of it are largely absent (Andreasen & Moller-Jensen, 2016; Foster et al., 2010; Lawhon et al., 2014; Obeng-Odoom, 2012).

(20)

12 However, centralised water infrastructure in many cities of the South has never served the majority of urban populations (e.g. Kjellén & McGranahan, 2006). Even where the network reaches, service is often intermittent (Burt & Ray 2014; McIntosh, 2014) and of non-potable quality (Onda, LoBuglio, & Bartram, 2012). Instead, urban residents in Southern cities rely on a combination of multiple water sources to fulfil their water needs (Ayalew et al., 2013; JMP, 2014). These include, for instance, groundwater (Chakava, Franceys, & Parker, 2014; Wright & Jacobs, 2016), rainwater (Nastar, 2014), or even ‘wastewater’ (Meehan, Ormerod, & Moore, 2013).

Likewise, water provided through both formal and informal small-scale private vendors is another important vehicle of access in many cities of the South (Kariuki & Schwartz, 2005). Some research on small-scale providers covers access to groundwater through shallow wells or boreholes (Ayalew et al., 2014; Kjellén & McGranahan, 2006). Nevertheless, most of the research considers those providers, who, formally or informally, source and resell water from the central network and function as an extension of it, such as standpipe and kiosk operators, or ‘handcart pushers’ (Albu & Nijru, 2002; Solo, 1999; van Dijk, 2006). While they have been praised for their functioning as a decentralised form of water provision, more often, small-scale local water vendors (like packaged water, see section 2.1) are seen as a ‘failure’ of the piped system, particularly in international development (e.g. ADB, 2013; Goldblatt, 1999). Criticised for high prices and uncertain water quality, water vendors are presented as a symptom of the disconnection of the poor from piped water, resulting in the poorest paying much higher prices for water, than connected households (ADB, 2013; WSP, 2015). Water vendors are thus used as an argument for the ‘pro-poor’ characteristics of piped water and likewise, the idea is that ‘fixing’ the piped network will abolish the need for alternative sources like water vending.

In contrast, critical perspectives on urban water are increasingly emphasising the normality of this diversity of water access configurations, and question traditional conceptualisations of water access based on universal, centralised-network access. They point out that these models might be unsuitable for Southern contexts and might even work against equity of access: as Robinson (2002) points ‘the colonial model of piping water to every homestead has ironically worked against the issue of equity. It was too expensive to replicate, so in the end, virtually nothing was done’ (Robinson, 2002, p. 852). Within a wider theoretical framework of Southern Urbanism, these scholars are challenging the applicability of urban theories rooted in the experiences of Northern cities to cities of the Global South, and call for an expansion and reframing of how ‘the urban’ is understood in contemporary urban theory (Parnell & Oldfield, 2014; Roy, 2011).

These perspectives allow for a critical reconsideration of ‘pro-poor’ urban water governance, which is important, since theories grounded in the infrastructural ideal of universal access to a centralised network has concrete implications for water-related development approaches: for instance, the complex configurations of urban water supply, as described above, were not ‘counted’ within the MDGs, as a result of their focus on

(21)

13 ‘improved’ sources in general, and piped water in particular (Nganyanyuka et al., 2014). Thereby, development approaches failing to expand their framework of analysis beyond piped water, are also at risk of failing to capture water access realities of the urban poor in cities of the Global South – a profound problem also identified in the MDGs (Satterthwaite, 2016). In contrast, studying water access through a lens of Southern Urbanism means reassessing this conventional model, as illustrated by the following quote:

‘I propose a radical change in perspective, taking as a starting point not the failure of urban services and the institutions responsible for their delivery, but the vitality and multiplicity of actual delivery systems which, despite policy announcements and reforms, and notwithstanding imported models, survive and contribute to the functioning of cities’ (Jaglin, 2014, p. 434).

By neglecting these ‘alternative’ modalities of water access, research on urban water has so far missed out on exploring their positive and negative potential for equity of water access. Equity is at the heart of an Inclusive Development approach and a substantial component of the water-related SDGs, which are described in the following section. It is particularly in this time, in which the developing community structures its ambitions around a new set of development goals, that it is necessary to re-assess the way access to water is analysed. Approaching this topic from the theoretical angle of Southern Urbanism can contribute to a more differentiated understanding of the Southern urban experience.

2.3 Inclusive Development and Equity of Access

Situating Pro-Poor Water Provision Within Inclusive Development

The ideas and ideals now enshrined in the SDGs have been part of international development discourses quite some time before their formal adoption by the United Nations General Assembly in September 2015. As the name indicates, the SDGs are a product and institutionalisation of a Sustainable-Development approach (Gupta et al., 2014). While Sustainable Development is most readily associated with environmental factors, it is now generally acknowledged that in order for development efforts to be sustainable, they need to also pay attention to components of social inclusion and economic development (Sachs, 2004; Sachs, 2012). However, multiple scholars have criticised Sustainable Development and the way it is implicated in practice (Dubash, 2012; Lele, 1988). These authors criticise that Sustainable Development frequently leads to trade-offs made among its dimensions in favour of its economic and environmental components and at the expense of social inclusion (Gupta et al., 2015a), a deficit which has also been identified in the SDGs (Gupta et al., 2014). Compromised by dominant neoliberal influences in development and an overwhelming focus on economic growth, this has led to a ‘weak’ version of Sustainable Development (Gupta, 2014).

One way to address this issue has led critics to frame their questions within an Inclusive Development framework. As of now, Inclusive Development remains an emerging

(22)

14 perspective, and only very recently have steps been taken to create a comprehensive theory of Inclusive Development (Gupta et al., 2015a; see Box 1 in Appendix 1). Taking an Inclusive Development perspective means re-centring concerns of inclusion and social equity in the development debate and thereby challenging the dominance of economic and ecological sustainability over social sustainability. Arguing along those lines, striving for inclusion means the elimination of exclusion. This resonates with the ideas of Human Development, most profoundly influenced by the work of Amartya Sen and his capabilities approach, which places human beings and their freedoms at the centre of development. Here, “[d]evelopment can be seen … as a process of expanding the real freedoms that people enjoy. Focusing on human freedoms contrasts with narrower views of development, such as identifying development with the growth of gross national product, or with the rise in personal incomes, or with industrialization, or with technological advance, or with social modernization” (Sen, 2000, p.3). Following Sen, Chatterjee (2005) defines Inclusive Development as “a development process that generates broad-based participation, and specifically reduces poverty and social exclusions” (p. 13).

Being able to access clean water is one such freedom, which also impacts other types of freedoms, such as being healthy, and enjoying well-being, while being restricted from water access presents a type of unfreedom. Therefore, a central ambition of Inclusive Development approaches is it to counteract marginalization (Von Braun & Gatzweiler, 2014), which entails a specific focus on the most vulnerable groups in society (Cook, 2006; Palomino-Schalscha et al., 2016; Sutherland et al., 2015). Thereby, Inclusive Development adds an additional emphasis on equity concerns in development. And indeed, one of the first publications using the concept is a 2007 Asian Development Bank paper, where Inclusive Development is used as a strategy for the achievement of equity through, among others, poverty reduction (Gupta et al., 2015, p. 549). In addition, there is a link between water access and poverty reduction: For instance, increasing water access has been shown to reduce poverty on the provincial level (Larsen, Pham & Rama, 2004).

As Chatterjee (2005) shows on the example of poverty reduction in Asia, ‘[t]he success of an Inclusive Development strategy can be gauged therefore by the extent to which such a strategy is able to reduce poverty and social exclusions’ (p. 13). However, as Kanbur & Rauniyar (2010) suggest, Inclusive Development can have varying degrees of ‘pro-poorness’, which depends on the extent to which it specifically targets poor populations (as opposed to other categories of exclusion, such as gender4_{). An inclusive strategy is} therefore necessarily a pro-poor strategy, and inclusion of ‘the poor’ supports the inclusive-development process. It is particularly for its explicit link between inclusiveness, equity, and pro-poor approaches that Inclusive Development is used for this research, which has particular implications for urban water governance. When Inclusive Development means re-centring concerns of social inclusion, we need to assess who is included and excluded from water access in cities of the Global South and why.

(23)

15 Existing literature identifies a two-way relationship between water access and inclusiveness: employing inclusive strategies to water provision can lead to an expansion of access to marginalised communities (Singh et al., 2009; Tortajada, 2014), such as the urban poor (Chatterjee, 2005). Likewise, inclusive water governance strategies have the potential to contribute significantly to increasing sustainability of the urban water sector (Singh et al., 2009). For instance, strengthening the role of women in urban water governance can contribute to increasing resilience and urban sustainability in the face of climate change (Figueiredo & Perkins, 2013). Thereby, Inclusive Development has made a full cycle back to Sustainable Development.

The SDGs, Water, and Inclusive Cities

Goal 6 of the SDGs is it to ‘ensure availability and sustainable management of water and sanitation for all’, while its first target (6.1). aims to ‘achieve universal and equitable access to safe and affordable drinking water for all [by 2030]’ (United Nations, 2016, see Box 2 in Appendix 1). By conceptualising ‘water access’ along four central dimensions beyond the existence of ‘improved’ water sources, Goal 6.1 therefore makes some important adjustments vis-à-vis its preceding MDG Target 7.C. and the sole focus on increasing access to ‘improved’ water sours. This expansion addresses the complexity of water access and responds to some of the main criticisms towards the MDGs, particularly through its attention to the quality of water accessed, and a focus on equity of access, a central component of pro-poor water access strategies within an Inclusive-Development framework (section 2.3.1).

Inclusiveness is also part of Goal 11: ‘Make cities and human settlements inclusive, safe, resilient and sustainable’ (United Nations, 2016). Both Goals are explicitly and necessarily linked: As Sutherland, Scott, and Hordijk (2015) point out, ‘[t]he provision of adequate water and sanitation is essential in creating a more inclusive city’ (p. 488). And indeed, the central importance of cities for development have been re-emphasised in the context of the SDGs as not just a site of implementation, but as their central drivers. As Parnell (2016) suggests: ‘cities are key pathways in every aspect of Sustainable Development and not merely vehicles for the promotion of social, economic, or environmental objectives’ (p. 539). Therefore, she calls upon the international development community to ‘spell out the centrality of urban hubs in the implementation of all of the SDGs’ (Parnell, 2016, p. 538).

In addition to Inclusive Development, approaches informed by Southern Urbanism are necessary in the context of the SDGs. At the time of writing, the main indicator proposed by the Inter-Agency and Expert Group on Sustainable Development Goal Indicators (IAEG-SDGs) for Goal 6.1 is the ‘percentage of population using safely managed drinking water services’ (IAEG-SDGs, April 2016). Upon closer examination, the indicator’s definition strongly resembles the one used for MDG Target 7.C, namely the ‘[p]opulation using a basic drinking water source (‘improved’ sources of drinking water used for MDG monitoring […]) which is located on premises and available when needed and free of faecal (and priority chemical) contamination’ (ibid.)

(24)

16 Recycling the focus on ‘improved’ sources places the SDGs at risk of once again failing to ‘count what counts’, despite adding more dimensions to the Goal itself. Having explained the value and importance of viewing water access from the perspective of Southern Urbanism, this research project aims to expand the conceptualisation of inclusive water access. In order to do so, it refers to SDG Goal 6.1’s four dimensions (universal, equitable, safe, affordable) as guidelines alongside which the role of refill water in Jakarta is evaluated, in order to assess in what way and under which circumstances refill water plays a role in facilitating or inhibiting the achievement of this goal. Chapter 4 therefore assesses what gap in water access refill water is filling and who it is providing access to (universal); Chapter 5 investigates the quality and price of refill water (safe, affordable); and Chapter 6 explores dimensions of equity and inclusiveness of refill water by showing how it is accessed differently by different socioeconomic groups (equitable). Thereby, it hopes to show how Inclusive Development, viewed from a perspective of Southern Urbanism, can contribute to capturing water access realities beyond the piped network as a basis for the implementation of the new Global Goals.

(25)

17

Methodology

This chapter maps out the methodological setup of this research project. Research questions, epistemological and ontological positions, and the mixed-methods research design are introduced, before proceeding to show how the data collection has been carried out. Finally, quality criteria of the collected data and ethical considerations are addressed.

3.1 Research Questions

This research project is structured around the following three research question, which incorporate the dimensions of SDG Goal 6.1.

Main Research Question: What is the role of refill water in providing inclusive access to water supply in Jakarta?

Sub-Questions:

RQ1: How does the refill water sector fit into the landscape of urban water provision in Jakarta? Who is refill water providing access to?

 What is the historical development of refill water, why, when and how did refill water come into existence?

 In what way can refill water be considered a response to inadequacy of other water sources in Jakarta?

 How and by whom is refill water accessed?

RQ2: What is the quality and affordability of this access?  How is refill water regulated in policy and in practice?

 What are the practices of vendors with regard to water safety and quality maintenance?

 How do customers, vendors, government, and donors perceive the price and safety of refill water?

RQ3: How does refill water meet goals of providing equitable access to water?  How is refill water accessed differently by different socioeconomic groups?  What is the environmental impact of refill water and can it be considered

sustainable?

 To what degree is refill water recognised by key actors and institutions in government and (inter)national development?

(26)

18

3.2 Epistemological and Ontological Position

Epistemology: Critical Realism

This research project was conceptualised and later executed following a critical realist epistemology. Critical realism has often been placed in the middle of the spectrum of epistemological positions, ranging from positivism on the one end to interpretivism on the other. Critical realists acknowledge that ‘objects (including other people’s ideas) [are independent] from our thoughts about them (Sayer, 2006, pp. 99-100). However, while accepting the positivist stance of the existence of an objective reality of the social world, critical realism acknowledges that this reality is influenced by the researcher – it can only be uncovered partially, because of the inherent subjectivity and complexity of this social world (Bryman, 2012, p. 29). As a consequence, adopting a critical realist perspective requires the researcher to pay attention to: a) the source of the construction of reality, b) the history and timespan of this construction and c) the fact that construction requires physical and non-physical materials (Sayer, 2006).

These considerations build the foundation of this research and informed the choice of a mixed-methods research design (see section 3.6). Following a critical realist epistemology, the goal is to understand, rather than to explain. In this framework, the role of refill water is a factual thing, which can be uncovered - at the same time, the complexity of the social world means one study only allows for a glimpse into what might be part of ‘the truth’. Incorporating abovementioned three principles of critical realism means including as many different perspectives as possible: of the people on the supply-side of refill water (spring management, transportation firms and the refill stations themselves), of the customers (presumably low-income households), and government officials, aid workers and NGO employees (a). These interviews set the foundation for the discovery of non-physical materials, like beliefs and habits with regard to drinking water consumption (c). In addition, secondary data facilitates the contextualisation of this research project within the fourteen years of refill water existence (b). Thereby, this research is set out to take a first step into untangling the reality of refill water and its implications for equity and inclusiveness in urban water access.

Ontology: Substantivist and Relational

In relation to this epistemology, this research adopts both substantivist and relational ontologies, depending on the unit of analysis. For instance, it aims to capture a snapshot of the status quo of drinking water access and refill water in a specific place and time and thereby usually corresponds with a substantivist ontology. At the same time, while ‘drinking water’ is understood as a clearly-defined, independent object that exists outside of social relationships, ‘access to drinking water’ is understood as an inherently negotiated and relational concept. This is also reflected in the choice of MMR as methodology, which aims to both quantify the role of refill water (through the customer survey) and to get an institutional overview over all the people, processes, and ‘materials’ (in the definition of Sayer, 2006) involved in what becomes the refill water market. Likewise, refill water does only exist in relation to other water sources – as ‘a consequence of’, ‘in spite of’, ‘alongside’, ‘instead of’. This relational interconnection

Pro-poor bottled water? : the role of refill water in providing inclusive access to water for low-income communities in Jakarta, Indonesia