Sites of entitlement: claim, negotiation and struggle in Mumbai

This paper develops a conception of “sites of entitlement” as a basis for better understanding how infrastructure and services are perceived and experienced in informal settlements. While legal and policy frameworks are often viewed as the source of entitlements to infrastructure and services, the complexity of provision, access and negotiation in informal settlements demands a conception of entitlement that exceeds those domains. Based on ethnographic research on sanitation and water in informal settlements in Mumbai, we focus on the ways in which people’s everyday experiences, interactions and practices constitute sites of entitlement. These sites are unevenly produced, contested, often in flux and ambivalent, sometimes made through collective struggle and at other times through quiet individual practice, and always constituted by social relations. Sites of entitlement emerge in close relation to moral economies, and are characterized by often profound and – for research, policy and practice – challenging levels of spatial and temporal variation. We argue that sites of entitlement are vital for thinking through the possibilities of realizing the universal right to sanitation and water.


I. INTRODUCTION
Increasing concerns over the consequences of inadequate urban sanitation and water for poverty, health, livelihoods, and education have spurred global declarations on the human right to sanitation and water throughout the past decade. (1) However, the social and spatial heterogeneity of urban poverty is often missing in global policy debates. (2) Sanitation interventions usually focus on informal settlements that are officially recognized, and there is also a generally weak understanding of the micropolitical realities of urban services in informal settlements over time and space. While research, policy and practice debates routinely invoke the human right to sanitation and water, there is comparatively little grasp of how local understandings of entitlement are produced and contested, or how that varies across social groups and spaces. There is a pressing need for greater attention to the geographies of entitlement in approaches to water and sanitation. In this paper, we offer a focus on what we are calling "sites of entitlement" as a basis to carry this approach forward. We make our case based on ethnographic fieldwork conducted in informal settlements in Mumbai.
Entitlements are usually considered at the level of state, legal and policy frameworks, and in relation to how these interact with global agreements and discourses such as the Millennium Development Goals. While the complexity of provision, access and negotiation around sanitation and water in informal settlements is often gestured to in these domains, it is also routinely demoted to secondary importanceto mere context to the abstractions of broader principles and agreements. Our research suggests a different view is required. The geographies of sanitation and water provision and access across and within urban informal settlements are profoundly diverse and contingent on various social, economic, political, institutional, technological and historical processes. We see this within one city, let alone between cities and countries. Research over the past decade has widened our understanding of these urban geographies and processes in significant ways. (3) This diversity and contingency should leave us in no doubt that the urban poor have strikingly inadequate formal rights. As Partha Chatterjee has so influentially shown, the fact that so many of the urban poor live in, at best, an ambiguous relationship with dominant frameworks of private property and citizenship rights leads to an often ad hoc, piecemeal provision of basic services and infrastructure across the global South. (4) Even when certain rights are granted, they are often not realized because of the uneven power relations that shape the everyday life of the city, and the poor very often turn to diverse informal arrangements, social networks and negotiations to meet their sanitation and water needs.
The argument we make is more than a gesture to spatial diversity. We believe this spatial diversity demands that we refute any easy separation of abstract principlesuch as the right to sanitation and waterand messy everyday contexts and struggles. More fundamentally, we position entitlement alongside this spatial heterogeneity, and on the same conceptual plane. Rather than seeing context as the "micro" and principles of entitlement as the "macro", we contend that it is impossible to think about one without the other. Universal rights to sanitation and water can only emerge through a focus on the everyday experiences, claims, negotiations and struggles that continually take place, sometimes through overt individual or collective action and sometimes through quiet processes of subversion, in informal settlements. This is what we mean by "sites of entitlement": the always coexisting principles, laws and norms with social and spatial differences in the everyday production of claims. Sites of entitlement are characterized by changes over time, often require ongoing negotiation, and are frequently characterized by uncertainty and ambivalence. Sites of entitlement are the lived geographies of abstract claims to universalist rights, fundamental to how rights to sanitation and water are delivered or might be delivered in the future.
We propose that "moral economies" play a significant role in producing sites of entitlement. We define moral economies as both collectively understood informal regulations around expected behaviour and an individually held sense of what is expected that may or may not coincide with that shared collective view. While moral economies are collectively shaped in the lives and local/translocal relations that people have, individuals often construct their own preferences and contestations in relation to those collective views, sometimes in oppositional ways, and this dialectic shapes a sense of what individuals feel they might reasonably expect from others. Moral economies provide a legitimate, though not necessarily legal, basis for claims and affect how entitlements are conceived, claimed, contested, and ultimately realized or denied, along with other factors such as local power relations, state action or regulation, and so on. Sites of entitlement are relationally formed between different groups, spaces, actorsfrom states to individual politicians to civil society groups or private actorsand moral economies.
Moral economies do not play a one-way causal role here. In practice, moral economies and sites of entitlement are constructed fluidly. Moral economies can delimit a provisional sense of who is and is not entitled to provisions and under what conditions, while a sense of entitlementwhether locally based on cultural historical practices or legally basedcan play a part in generating the operative moral economy. The relation between the two emerges more as a contingent assemblage than a linear causal process. Sites of entitlement reveal a complex and contingent relation between moral economies and entitlement. For example, while some people, and in relation to some services, might feel a strong sense of moral economy around state delivery, experience may indicate that such provision is so unlikely that to think of it as an entitlement borders on absurd. Moreover, this relation between moral economy and entitlement may vary from group to group or neighbourhood to neighbourhood. Moral economies are important to shaping entitlement, but not in a linear way. We argue that our focus on the relations between everyday life, moral economies, and sites of entitlement lays out an important and poorly understood research and policy agenda. From this we can develop a more nuanced understanding of the challenges and, through this, a more socially equitable provision of basic services.
The paper is based on a nine-month ethnographic study in two informal settlements in Mumbai. Rafinagar is a poor unauthorized or "non-notified" settlement in the north-east with an older, more established Part 1 and a newer, poorer and still expanding Part 2. Khotwadi is an authorized, well-established settlement in the north-west. (5) Khotwadi, with a population of approximately 2,000 households, has 24 toilet blocks and a total of 180 seats, whereas Rafinagar, with approximately 4,000 households, has six toilet blocks and 76 seats. Rafinagar, then, has twice the population and half the toilet seats. There is no legal requirement for the state to provide services and infrastructure to non-notified Rafinagar, but all informal neighbourhoodsnotified and non-notifiedcan access services through specific projects. One such project is the Slum Sanitation Programme (initially supported by the World Bank in the mid-1990s, which provided toilet blocks to Rafinagar (6) ). These are unevenly distributed: Rafinagar Part 2 has only one toilet block, provided by the Maharashtra Housing and Area Development Authority (MHADA) in 2011, although also in use are temporary hanging latrinesprecarious self-built structures usually over open drains or watercourses. While most residents in Khotwadi have a degree of secure water access through unmetered municipal standposts, metered group connections and wells, most of Rafinagar's residents face profound difficulties and incur high expenditures for water and/or time and effort in collecting water. The research involved participant observation of different areas in both neighbourhoods, extensive and repeated interviews with a cross-section of people from both neighbourhoods, and interviews with municipal staff, non-governmental organizations, and community groups working on sanitation and water.
The production of local conceptions and practices of entitlement in the two neighbourhoods plays out in quite different ways. It is closely related to a sense of moral economy in relation to the everyday state, private sanitation providers, and the exclusion of particular groups. Our argument is that sites of entitlement emerge in large part through a series of contextually specific moral economies and by linking a wide variety of everyday practices across different spaces. These particular issues are those that matter most in the informal settlements where we conducted the research. Our larger point is, however, precisely to underline this spatial contingency as a more active, vital basis for thinking about universalist claims to the right to sanitation and water, rather than as an afterthought to those claims.

II. COMPOSING EVERYDAY ENTITLEMENTS
We begin by asking a seemingly straightforward question: how are people's sanitation and water entitlements produced? To consider this question, we must first clarify what we mean by entitlement. The language of entitlement is widely used in two ways: to invoke universal normative rights, or in relation to specific rights attached to certain groups. It is important to distinguish here between "rights" and "entitlement". While they are interrelated, we take entitlements to be what people claim as a product of the rights available to them. Importantly, while rights generally take the form of legally binding statements, entitlements are produced through social relations and based on people's experience and perceptions. Amartya Sen's influential writings are useful here. For Sen, a person's entitlements are the totality of thingsresources, commodities, infrastructures, services, and so onthat she can command based on all the rights applicable to her. (7) Sen argued that while it is usual to characterize rights as a relation between particular actors, such as people and the state, entitlement emerges by virtue of provision of rights. (8) If Sen tends to conceptualize the basis of these claimsthat is, the source of their legitimacy in narrowly legalistic terms, some of his work on gender and poverty does attempt to account for the role of social conventions that govern the basis of claims beyond courts. (9) A number of scholars have sought to expand the concept of entitlement to include the things an individual can command through claims with a basis in other sources of legitimacy. Charles Gore, (10) for instance, has argued that entitlements are shaped not only by legal rules but also by socially shared rules. Naila Kabeer (11) has called for a reformulation of Sen's concept of entitlement to consider the different basis of claims on resources that prevail in a society, and that are embedded within social relations and practices rather than legally binding contracts. She argued that entitlements are generated through rules, norms and practices characterizing different institutional arenassuch as market-based exchange, state provision, community and kinshipthat determine who gets what and on what terms. For Gore, these socially shared rules are best understood through the concept of "moral economy". There is a long history to this idea, of course, and its use in the social sciences owes a great deal to James Scott's 1976 book The Moral Economy of the Peasant, which emphasized the role of shared traditional methods and models put to work by rural communities in Southeast Asia to cope with change. (12) Broadly speaking, moral economies are social regulations related to but not determined by legal rules, which help set the conditions through which entitlements are conceived, claimed and contested. Moreover, they vary within and between different groups and spaces, and over time. As Gore has argued, (13) although the concept of a moral economy refers to socially shared rules, based, for instance, on moral obligations, patron-client networks and so on, senses of entitlement are not necessarily shared by all and need to be seen as an outcome of struggles and active processes of negotiation.
We conceptualize entitlement as a claim that relates a person/persons to services/infrastructures (water, drainage, toilets, etc.) in a way that seeks to fulfil their access to what they need and value, and which is produced through diverse practices, social relations, and moral economies. While we value the emphasis by scholars such as Scott, Gore and Kabeer on nonlegalistic claims to entitlement through moral economies, we do not differentiate between socially shared rules articulated through a moral economy and legal rules, since the latter are often an outcome of a particular moral economy and can also feed into moral economies. Rather, we emphasize the importance of unpacking the moral economies, social relations and practices through which individuals or groups both shape sites of entitlement and are positioned as deserving or undeserving of entitlement. As studies on informality, urban politics and citizenship reveal, it is through individual and collective practices of negotiating urban space and resources that marginalized groups attempt to expand their claims. (14) While active collective mobilization is important in these accounts, this body of literature suggests that entitlements are produced through different modalities of urban struggle and people's everyday experiences and practices.
We consider it vital that research on entitlements focus on the production of entitlements from the perspective of residents themselvesin this case in relation to informal settlementsand from the perspective of their moral economies. Quite apart from the centrality of moral economies to people's sense of entitlement, there are two other important reasons for focusing on sites of entitlement.
First, the entitlement of residents is too often understood through a state-centric lens. This leads to an overly legalistic, formal view of entitlements. We are not arguing that the state is anything but critical to how residents understand entitlement, nor are we suggesting that the state exists "above" the daily life of the neighbourhoods. The state, in the form of local officials, ward councils, specific projects, or particular politicians, is present in specific ways in the daily life of the neighbourhoods and in shaping a sense of entitlement. Instead, we argue that rather than focusing on the state's treatment of particular informal settlements as a basis for understanding entitlement, it is more fruitful to examine the currents of everyday life and why and how the state features in people's understandings of entitlement. As research on the state has shown, this approach typically reveals a highly variegated sense of how the state shapes local perceptions and practices. Ethnographic studies reveal the state not as an entity "up there", above society, but as constituted by a multiplicity of actors, sites and practices that permeate and are reshaped through society and everyday life. (15) Residents of informal settlements encounter and experience the state through the practices of elected representatives, municipal officials, the police and others in a variety of modalities, whether violent, oppressive, beneficial, welfarist or governmental. These relations influence and are influenced by the moral economies of these different actors.
While the state-centric focus on entitlement has been enlarged in recent decades to account for the increasing importance of frameworks of institutional pluralism, and especially the state's role in facilitating service delivery through civil society and the private sector, the everyday experience and perceptions of communities around entitlement remains largely neglected. As non-governmental organizations (NGOs), community-based organizations (CBOs) and the private sector have increasingly become service providers or partners in service provisionincluding in sanitation in Mumbai (16) the shaping of entitlement is more and more understood as a relational process in which the state sets vital parameters. We would not of course deny the vital role these actors play in shaping perceptions of entitlement, but too often the claims and views of established organizationsespecially states, NGOs and social movementsare taken as representative of people's entitlements. As a result, social relations of family, kin, community and locality are often removed from view. Rather than look to mediating organizations as representatives of people's sense of entitlement, a sites of entitlement approach aims to empirically examine the different social relationsand civil society groups and states are usually and inevitably part of thisthat compose conceptions and practices of entitlement.
Second, a sites of entitlement approach focuses on how residents compose a sense of entitlement not just across space, but over time. As we shall see, this reveals senses of entitlement in variable amounts of flux and ambivalence. It also reveals how these understandings and the moral economies that help shape them become contested, undermined, or altered through new relations, including changing political relations. Of course, not all practices to improve access to sanitation and water are about entitlements per se. Residents in Rafinagar and Khotwadi do not always have a clearly articulated sense of entitlement that informs their practices. Often, their practices are about fulfilling their needs in conditions of profound uncertainty and inequality. However, when established practices to fulfil needs are under threat, residents express a sense of injustice and, alongside this, a sense of their legitimate claims.

III. MORAL ECONOMIES, LEGALITY AND THE STATE
What, then, is the role of the state in producing sites of entitlements? We begin in Rafinagar, with a key issue that shapes but, as we shall see, does not determine, the nature of understandings of local entitlement: legality and the state.
Salimbhai is an elderly shopkeeper who has owned a furniture workshop in Rafinagar since 1985. He explained that demolitions had been common before he came to Rafinagar. Along with some other residents and shopkeepers, he established a committee and succeeded in obtaining the support of a local politician -Javed Khanto prevent further demolitions. Khan was a Member of Legislative Assembly (MLA) (17) for the constituency between 1985 and 1995. Salimbhai explained that the committee had tried to get Rafinagar declared a notified slum (18) with Khan's support but had been unsuccessful. But Khan had supported the committee in other ways and Rafinagar got its first water connections, streetlights and toilet blocks as a result in the late 1980s, more than a decade after its emergence. Salimbhai argued that if Khan's reputation had not been damaged by involvement in the 1992 Hindu-Muslim riots, he would have been impossible to defeat in the area because of his good work.
An informal settlement must be notified (declared) as a slum under the Maharashtra Slum Areas (Improvement, Clearance and Redevelopment) Act, 1971, for it to legally qualify for the provision of basic amenities such as water, toilets, streetlights and drainage. Rafinagar's experience shows, however, that despite no legal recognition, the relations with Javed Khan initiated the cultivation of relations with local elected representatives and a kind of extra-legal recognition. Residents were thus able to slowly carve out claims through negotiations in political societya disparate realm largely outside of formal citizenship claims in which community groups make moral claims, supplementing electoral relations, on particular politicians and officials. (19) However, political society has brought mixed results, such as they are, to Rafinagar. Negotiations have led to some successes in gaining infrastructure and services in Rafinagar Part 1, the older and more established area of the settlement, but have been highly uneven. For example, water cannot be legally provided to residents who arrived in the neighbourhood after 1 January 1995, meaning that these residents are doubly excluded through both the legality of the space and the time of inhabitation (pre-1995). Residents estimate that only 1,500 out of a total of 4,000 households in Rafinagar can prove pre-1995 residency. (20) Later we will consider the unsuccessful attempts by residents in Part 2, the area settled after 1996, to negotiate entitlements through political society.
The distribution of entitlements has also been significantly shaped by each family's personal circumstances and its networks and negotiations with elected representatives. Consider the experience of Shobha, one of Khotwadi's poorer residents. Even though Khotwadi is a notified informal settlement and Shobha and her neighbours are pre-1995 residents, obtaining a shared water connection as per the policy of the Municipal Corporation of Greater Mumbai (MCGM or BMC) has been a fraught process. Encounters with unresponsive municipal officials often force residents, even when they are legally entitled to a shared water connection, to approach their representatives for what can also be a long, frustrating process. Even after Shobha and other households had each paid the municipal corporator Rs. 1,200 (US$ 19 as of March 2015) for a shared water connection, the infrastructure had yet to materialize. Shobha said she was now worried that if the municipal corporator changed she would have to begin the whole process again, adding: "The government should do something about this. I can't spend much money on this. We have already paid Rs. 1,200 each. How much more will we have to pay who knows? If [the money] is adequate then he will give us a connection. It might work for two years, four years…but a sarkari [government] tap is a sarkari tap, right?".
Shobha's narrative of her ongoing attempt to get a water connection reveals that even when legally entitled to certain infrastructure and services, residents are often forced to negotiate in political society to realize these entitlements in practice, an experience that can be a source of immense anxiety and uncertainty. Shobha argued that the state should take responsibility towards residents like herself and provide sarkari taps, i.e. unmetered municipal standposts, in her neighbourhood. Khotwadi has often fared better in water provision than other localities in part because it is predominantly Marathi. Therefore it is favoured by the dominant political party in the city, the Shiv Sena, a nativist and historically anti-Muslim party that has run the BMC since 1995. (21) However, provision has been uneven: the Shiv Sena municipal corporator extended the network of unmetered standposts in some of Khotwadi's predominantly Marathi neighbourhoods by taking a one-time fee from some residents, even as elsewhere in the same locality such standposts were removed by the BMC and replaced by water connections that entail monthly bills. What is striking about Shobha's account is that she expects the state to provide despite these spatiotemporal variations. Here, a moral economy of state provision remains, uncertain, frustrating and stuttering though it may be. Such an expectation of the state is less common in Rafinagar, where for some a sense of entitlement from an often hostile state is an alien notion.
Indeed, even through the medium of political society, the opportunities for provision from the state in Rafinagar have dwindled rather than expanded over the years. Many older residents argue that water problems have increased and intensified since Javed Khan's time. Neither elected representatives nor municipal officials have taken serious steps to resolve this water crisis. Instead, the BMC, with the police, carried out raids on water connections in Rafinagar, Shivajinagar (the larger area adjoining Rafinagar, which was first established as an official slum resettlement site) and other nearby areas in 2009 in the name of cracking down on "water theft" to deal with Mumbai's "water shortage". (22) Both legal and illegal connections were cut during the raids without adequate alternative provision, destabilizing provision previously gained, and Rafinagar was thrown into an even deeper water crisis than before. This is not a context where many feel that they can reliably make claims on state provision. Nonetheless, many residents have a strong moral economy of state provision of water, not so much for unmetered standposts as Shobha has in Khotwadi, but for water connections entailing monthly bills. The history of state provision and denial thus influences the moral economies of residents as well.
The Shiv Sena and the Maharashtra Navnirman Sena (MNSa party formed in 2006 by a faction that broke away from the Shiv Sena) have often run campaigns against north Indians and Muslims in the city. This chauvinistic ethnic and religious identity politics has influenced the workings of municipal government and shaped the hostile attitudes of municipal officials towards areas like Rafinagar and Shivajinagar whose residents are predominantly Muslim and sometimes north Indian. Moreover, municipal officials often see these areas as harbouring illegal Bangladeshi immigrants, despite the lack of clear evidence. The claims of residents, in particular certain ethnic and religious groups amongst them, to urban space and resources are increasingly delegitimized. This is particularly visible at moments such as that of "water shortage".
Many residents attempted to prove the legality of their connections that had been cut in the raids and thus obtain BMC permission to re-join their taps to the water mains; some succeeded, others did not. Residents also repeatedly directed collective appeals and protests at their elected representatives. This led the BMC to install two temporary water storage tanks near the entrance to Rafinagar. These were to be filled by BMC water tankers free of cost to residents. However, tankers came irregularly, sometimes once a day, sometimes four times. Residents had to frequently approach their elected representatives to ask them to pressure the BMC to send more water tankers. Residents who lived further from the entrance pressured elected representatives to install more temporary tanks near their homes, which eventually led to the installation of five tanks over six months. Sites of entitlement are constantly unfolding across spatial and temporal configurations, and are continually negotiated through political society.
Similar stories emerge around sanitation. Most of Rafinagar's residents explained that Bullet Patil, the well-known former municipal corporator, did not respond to sanitation complaints unless they were prepared to repeatedly plead for help at his office. Many residents had given up approaching him altogether and instead pooled money or their own labour to clean their drains or construct makeshift toilets. Given this, it is not surprising that some residents did not even consider the municipal cleaning of drains as a service they were entitled to. Some residents had a clearly defined spatial imaginary of entitlement here, differentiating between the smaller drains in their lanes that they felt responsible for and the larger drains on the main streets that they felt the BMC was responsible for. In contrast, in Khotwadi the cleaning of all drains has long been a relatively regular municipal service, and residents readily described it as an entitlement they expect from the state.
However, it is not the case that residents in Rafinagar do not believe in state provision. For example, three attempts had been made over the past few years to demolish a public toilet block in Rafinagar and replace it with a private blockevery time, residents came together to successfully protest and drive away the demolition crew. Residents cited their concern about having to pay more for a private block; some also explained that the three other private blocks in Rafinagar remained closed at night from 12-5 am, and conversion of this block would also mean no night provision. One resident explained that one private block was locked in the middle of the day if there was inadequate water from tankers. Residents do not necessarily feel a sense of entitlement to state-provided toiletssome said they did, others had less of an expectationbut they have a keen sense that state provision is preferable. In other words, while there is a strong moral economy of public sanitation in Rafinagar, as a result of bitter and often violent experience of demolition this does not necessarily translate into a clear sense of entitlement from the state. In sum, the role of the state in shaping sites of entitlement varies by place (within and between Rafinagar and Khotwadi) and issue (types of drains or toilets). If we delve deeper into the everyday practices and perceptions of residents in the neighbourhoods, we see that sites of entitlement are shaped by changing conditions and shifting moral economies of service and infrastructure that extend beyond the state.

IV. SHAPING ENTITLEMENT BEYOND THE STATE
The division between public and private is often blurred. Two of Rafinagar Part 1's three private toilet blocks were made possibledespite its "non-notified" statusthrough the support of elected representatives to the private organizations that built and operate them. Some residents who were earlier part of these private organizations were also proactive. Like most private services, these exceed but are not straightforwardly "separate" from the state. Nonetheless, they are first and foremost private blocks, and residents do not expect them to operate as public blocks. This does not mean that residents do not have compelling notions about what constitutes fair access to privatized sanitation. The sense of entitlement to private blocks is distinct, shaped through individual and collective negotiations around specific concerns such as user charges and water provision.
Consider one of the private toilet blocks in Rafinagar Part 1 where the operator attempted to increase the user charge from Rs. 1 to Rs. 2 per use some years ago. Naina, arguing that the toilet operator had increased the charge to increase his income, explained the reaction of resident users: The public created a scene. They went and sat down [to defecate] anywhere, in the maidan [open ground], the garden, on the road, near the clinic…So that he [the toilet block caretaker] will also not be able to sit there, he will also get the stink, no? Tomorrow he can even say it is Rs. 3. Should we drink water from his hands?... Meaning if he keeps increasing the money, should the public keep giving? And at that time if anyone came and said why are you sitting [to defecate] outside, then people just said go and tell him, he has made it Rs. 2. Then the area became dirtier…We told others. This way people stopped going there.
Naina explained that ultimately the toilet block operator had to decrease the charge to Rs. 1 again. Her narrative suggests that many residents felt that the increase was unaffordable and unjust, and that their individual actions linked up to register a protest against the increase. More recently, another private toilet block in Rafinagar increased its user charges to Rs. 2. Mumtaz explained what she thought about this: "Rs. 1 is enough to keep it clean. Why do they need to take Rs. 2?...Maybe because there is a water problem…But women complained so they take Rs. 1 from us…Women will not part with more money for the toilet. She will think of what she can get for her children instead." These are not transformative protests, of course, but they are part of a wider set of small acts through which a moral economy of sanitation is defended and a particular sense of entitlementnot to free private toilet blocks, but to reasonable chargesis maintained.
Mumtaz's sense of being entitled to access the toilet block for Rs. 1 and her rejection of the operator's attempt to charge Rs. 2 emerged then from the competing priorities that had to be negotiated in everyday life. This led to negotiations with the caretaker, who relented and allowed the women to use the block for Rs. 1. Some refused to pay the cost of using the toilet, or paid less. This was a gendered practice. Many men were observed entering the block after giving the caretaker Rs. 1 or even half a rupee. At times, the caretaker tried to stop them from entering, but often he seemed resigned, muttering once that men just paid what they wanted to. He clearly did not think this was okay but allowed it because he felt helpless to do anything about it. He said the toilet block was difficult to maintain because "here there are more taporis", a term for the figure of the rowdy male who is difficult to control. There was, then, a distinct gendered dimension to the caretaker's attitudes, and a distinct sense of entitlement across different groups of women and men. Other differentials and expectations occur: when women bring their young children to the block with them, the caretaker does not always charge for the child. In the caretaker's moral economy, women's claims to using the block for less than Rs. 2 were seen as more legitimate than men's, but men were more likely to ignore the caretaker's moral economy.
Sites of entitlement are, then, relationally formed through interactions among men and women, caretakers, and the nature of the toilet block (public or private). Moreover, sites of entitlement change over time as circumstances alter. Take the example of a toilet block built under the Slum Sanitation Programme (SSP) and run through a community-based organization in Khotwadi. Despite it being community managed, most users saw it as a private toilet. Before our research began, the CBO had closed off all piped water connections to prevent "water wastage" caused, according to the CBO, by women leaving the water tap running too long or not switching it off. Women users were then required to bring their own water from home or buy water from the caretaker: Rs. 1 for half a bucket of water and Rs. 2 for a full bucket. Shobha did not mind that the water had been closed off, and sheas well as other women users -echoed the CBO's narrative of "water wastage". But the caretaker also refused to provide water to clean the toilet if it was "dirty" when Shobha arrived to use it, and she resented using her own water to clean up after others. She added, further, that this response was class-based: "If there is a known person, a well-off person, they give water to them, but not to us. They see who is it and then give accordingly." In this example, Shobha's moral economy of water is shaped by both a local discourse of "women wasting water" and her questioning of the CBO's access practices. She saw it to be the CBO's responsibility to at least provide her with water at no cost if the toilet had been left dirty by a previous user, and she believed that all users were entitled to equal treatment. She raised these issues with the caretaker, to little avail, but not with the CBO itself: "Who would listen to one person?" she asked. In other words, while a moral economy of service provision helps Shobha to shape a sense of entitlement to a certain standard of provision, she had little faith that this could be realized in practice. Partly as a result, from time to time conflicts erupt between residents around who is entitled and who is not, and under what circumstances.

V. NEGOTIATING EXCLUSIONARY ENTITLEMENTS
In a context of inadequate sanitation facilities, conflicts sometimes erupted amongst residents over the use of toilet blocks in both Rafinagar and Khotwadi. Large numbers of users and the lack of water provision at the blocks often led to unclean toilet conditions, with residents often viewing particular groups as responsible for making the toilets dirty. In Khotwadi in particular, this led to discourses and practices through which residents sought to establish their entitlements by delegitimizing the claims of others, especially daytime garment workers from Khotwadi's many textile workshops.
Many public toilets in Khotwadi have been appropriated by particular groups of residents who regulate access to them through lock-and-key arrangements. Residents articulated the legitimacy of these appropriations by arguing that their own well-being was threatened. Lakshmi explained: "Illnesses spread through the toilet. Jaundice, diarrhoea and other diseases spread like this. The workers of the garment workshops make the toilets dirty. This is why we have put locks in. Women garment workers throw cloth into the toilet. This is why we do not allow them [to use the toilet]." In this and statements like it, residents demonstrated a moral economy of access that defends "their" resource, and underpins a sense of entitlement that links cleanliness to social identity.
In some blocks, appropriations were sometimes buttressed by historical territorial claims. A number of residents explained that the toilets in their chawls (neighbourhoods) had been built 40-50 years earlier by the informal landlords who built the chawl and rented out the units. Many toilet blocks in Khotwadi had thus been originally built for the tenants. However, the blocks had been rebuilt 25-30 years before by the BMC to convert them from dabba sandaas (dry toilets where night soil, or human excrement, had to be emptied manually) to pour flush toilets. As municipal funds had been used, it was no longer clear whether these historical territorial claims could be sustained. Prabha, for instance, was aware of this history, but considered it the "right" of residents of her chawl and some residents from an adjoining chawl to lock the toilet cubicles to prevent garment workers from using them. After all, they contributed money for the cleaning of the block and also for repairs and maintenance, so why should it not be appropriated more formally? She explained that residents in fact did not pressure the municipality to clean it, for then the toilet would be understood as public, which would create difficulties for the chawl residents: "If the BMC cleans it then everyone can go, so we don't force the BMC. If BMC cleans it then everyone will have a right (hakk). If we do it ourselves then it will remain private and personal." Here, then, historical ambivalence and poor municipal performance combine in the production of a particular moral economy that justifies semi-private control over a private-turned-public toilet, and produces a fragmented sense of entitlement that asserts the right of particular groups over other less powerful groups.
But this was not always the end of the story. Locks on these appropriated blocks were often broken, whether by garment workers or others, and in some cases residents relented by leaving open one cubicle for public use while the others were kept locked. Even then, locks were sometimes broken, and in one case residents had been forced to abandon the locks altogether. This does not mean that the claims of garment workers to the toilets were recognized by residents. As one garment worker explained, women residents from the area often shouted at them, blaming them for the unclean toilets. She added that one garment worker had even been slapped by a woman and was now scared of even going to the toilet. Although this did not stop her and other garment workers from using the block, verbal abuse and violence were sometimes used to attempt to establish territorial rights to the toilet blocks in the face of inadequate sanitation and the daily experiences this created for different groups.
For some, the issue of entitlement was not just one of controlling access to blocks but of charging at all. Shobha again: "We should not be charged. We are from here only, right? You can take money from outsiders, from the garment workers. Not from us. Why should you take from us? It should be a sarkari [public] toilet. [The CBO running the block] is also earning so much… from the garment workers. From one person they take Rs. 2 for a bucket of water. It was Rs. 1, now they take Rs. 2…They take Rs. 10-15 for bathing. So they must be earning so much." Sites of entitlement, then, are produced not just through moral economies rooted in notions of the public and the private, but through ideas about who should and should not gain access to these provisions, and about what constitutes acceptable forms of formal and informal regulation.

VI. CONCLUSIONS
Sites of entitlement are not fixed. They are ongoing processes that are reshaped through changing interactions between residents, between residents and states, between residents and civil society and private actors, and through the experience of living in often uncertain, precarious neighbourhoods. Sites of entitlement emerge not through one-way causal relations triggered by particular events, legal conditions or moral economies, but as contingent assemblages in which these different moments and processes shape one another in fluid and sometimes provisional ways. While we entirely support universalist claims to sanitation and water, our research leads us to argue that the understanding and production of entitlements in practice demand that research, policy and practice pay far more heed to spatial and temporal heterogeneity. Understandings of entitlement, like the provision of services and infrastructure themselves, are not just translated into local contexts; they are actively made and remade through spatial and temporal processes entangling multiple sites and actors.
The consequence for global debates on sanitation and water policy and practice is in our view significant: any effort to understand the extent and nature of the challenge and the process of working through solutions must engage with the heterogeneous nature of sites of entitlement not just between cities, but within them too. We are not suggesting that "quick fixes" or engineering-based approaches are no longer relevant, but our evidence shows that the success of any intervention depends on a strong understanding of people's lives, expectations and preferences, andcruciallythe ways in which they vary within and between spaces. (23) To be clear: we do not think it sufficient to continue with a gesture to how "local contexts matter": instead, we argue that the shift in focus from abstract entitlement to sites of entitlement holds the seeds of genuine success in understanding and addressing sanitation inadequacies. We need invocations of legal rights, state provision and community mobilization, but we need equally to recognize that such invocations often do not take us far in understanding the problems or working to solutions. Working in and with sites of entitlement at the neighbourhood level and across multiple actors is in our view the most effective route to lasting sanitation and water improvement, even though it is slow and particularist.
We end by highlighting two challenges for future research on sites of entitlement. First, the relation among legality, rights, moral economies, entitlement and everyday life is fundamental to both people's experiences and perceptions of water and sanitation and the possibilities of developing solutions. Working through solutions that assist better sanitation and water in Rafinagar must by necessity be distinct from those in Khotwadi. In Rafinagar, illegality is central, as is a more widespread politics of religion, and as is the politics of state hostility, manifested for instance in the profound challenges with water. In Khotwadi, the state is far more supportive and central to service delivery, including in relation to water, and the context for intervention is as a result radically different. These relations can and will change of course: a key part of the research challenge is to understand how the problems and solutions shift over time as conditions change.
Second, our research has demonstrated the need for better comparative understandings of basic urban services in informal settlements, not only between cities but within them. The fact that central aspects of the sanitation and water problems in Rafinagar and Khotwadi vary so significantly, including the production, logics, actors and form of the problems and potential solutions, reflects the value of comparative research. Comparative research at the neighbourhood level is particularly valuable here because neighbourhoods generally allow for a manageable unit of analysis between the multiple complexities of everyday life and the city more generally. Research, policy and practice tend to assume that spatial differences will be more pronounced between cities, and especially between countries, than within cities. The extent of the differences between Rafinagar and Khotwadi leads us to insert a note of caution here. Spatial diversity may be as strong within cities as it is between them, and we may find that particular neighbourhoods across different cities in different countries provide challenges in sanitation and water that have more in common with one another than with other neighbourhoods in the same city.

BIOGRAPHY
Colin McFarlane is an urban geographer whose work focuses on the experience and politics of informal neighbourhoods. This has involved research into the relations among informality, infrastructure and knowledge in urban India and elsewhere. A key part of this has been a focus on the experience and politics of sanitation in informal settlements in Mumbai, including the everyday cultures and contested politics of sanitation and water in two informal settlements. His current work examines the politicisation of informal neighbourhoods in comparative perspective, including African and South Asian cities. He is author of Learning the City: Knowledge and Translocal Assemblage (Blackwell), a book focused on the intersections among urban inequality, materiality, resistance and learning, as well as related edited collections, including: Infrastructural Lives: Urban Infrastructure in Context (Earthscan-Routledge, with Steve Graham), Urban Navigations: Politics, Space and the City in South Asia (Routledge, with Jonathan Anjaria), and Urban Informalities: Reflections on the Formal and Informal (Ashgate, with Michael Waibel).