Making space for women in urban governance? Leadership and claims-making in a Kerala slum

Abstract This paper looks at the role of gender in the shaping and exercise of political authority. Its empirical focus is a slum in central Trivandrum, Kerala's capital city, which is undergoing a phased process of formalisation and rebuilding funded through a flagship Indian national programme, the JNNURM. The upgrade project should offer a dense network of â€˜invitedâ€™ spaces for female participation within urban governance, both through women's presence within democratically elected municipal councils, and the deliberate linking of its implementation to Kudumbashree, Kerala's network of women-only neighbourhood groups that are responsible for implementing various antipoverty interventions throughout the state. Drawing on oral histories of the slum's evolution, interviews with project participants, and detailed ethnographic observation, we highlight the contests over identifying the list of JNNURM beneficiaries who would ultimately be granted a government-built flat at the project's completion. This key task in the project's implementation has been devolved to the local level, and therefore offers important insights into the practical efficacy of these invited spaces. The contests over this list show how â€˜actually existingâ€™ urban governance unfolds, and in particular highlight the interplay of formal and informal practices at work in â€˜fixingâ€™ a list that had local legitimacy. They also illustrate the ways in which power and authority are contested, and the role gender plays within performances of leadership. Women's political agency and efficacy are hampered both by Kerala's restrictive gender norms and by the high stakes and highly masculinist struggles present within its urban politics. The paper's theoretical contribution is to broaden our conceptualisation of leadership and claims making in the Global South, and within this to pay proper attention to the gendered nature of political space.

redevelopment. (2) Inevitably, this is a difficult point in the slum's evolution, where significant external resources enter the slum, permanently changing its physical form and people's rights to housing within it. It is also a point at which patterns of authority within the community are reshaped, as those aiming to control the change clash, and key decisions are made as to who will be included and excluded as beneficiaries of the progamme. Tracing the resolution of these conflicts therefore provides an excellent opportunity to view actually existing governance practices, and patterns of power and authority, within Kulamnagar.
In Kerala, women could be expected to have good opportunities to exercise and develop leadership roles through these struggles over programme implementation. Kudumbashree, its female-centred antipoverty mission, has directly involved poor women within formal processes of planning (Devika and Thampi, 2012;Devika et al, 2008;Williams and Thampi, 2013). This and a prior programme of decentralisation make the state's "institutional surface area" (Heller, 2009, page 85) relatively broad in Kerala: its ordinary citizens have a number of "invited spaces" (Cornwall and Coehlo, 2007) for participation in local governance, within which institutions have been deliberately crafted to ensure women's active presence. At the same time, the very nature of the upgrade programme poses challenges even for Kerala's welldeveloped governance structures. Its scale and the time pressures for its delivery, along with the complexity of the community it interacts with, make informal institutions and improvised governance practices a necessary part of programme implementation. This paper's key question is how actually existing political authority is shaped by gendered identities and practices: how is gender imbedded within this interface of formal and informal processes, what resources do male and female political leaders draw on in attempts to make binding decisions over the programme, and what limitations to their authority emerge as a result?
We address these questions by looking first at the literature on authority and claims making, and the (currently undertheorised) role of gender within this. Next, we review the slum upgrade programme, and the official processes through which it was intended to operate, before looking in detail at programme implementation between January and May 2013. The contests this sparked, and the differing roles of male and female protagonists within them, uncover the ways in which 'improvised' governance practices reproduced highly gendered ideas about local leadership and its legitimacy. Finally, we turn to the implications of this empirical work for the wider theorisation of political authority. Here, we argue that, although informality and 'bricolage' are necessary parts of governance-as-practised, attention needs to be given to their practical effects. Whilst they allow the construction of political authority beyond the roles envisaged within 'invited' spaces, they are highly ambivalent in producing leadership opportunities realisable by women.

Understanding political authority: gendering leadership and claims making
We begin from the premise that formal mechanisms of governance are necessarily incomplete and partial within planned development, particularly when, as in this slum upgrade programme, the state is intervening within a dynamic community in which paralegal sources of livelihood and forms of tenure are widespread. The literature on urban governance in India has frequently noted that the state itself exhibits partiality and irregular practices that extend from the unequal enforcement of its own rules [see Truelove and Mawdsley (2011) on policing Delhi's water use] through to the deliberate 'unmapping' of land use at the edge of expanding cities (Roy, 2002;. It is argued that through these practices, which often explicitly target informal settlements and their inhabitants (Baviskar, 2003;Ghertner, 2011), the informality (2) The JNNURM is currently active across sixty-five Indian cities, and offers significant central government subsidies for investment in urban infrastructure, which are themselves conditional on the implementation of programmes of urban reform: for reviews of the programme, see Shivaramakrishnan (2011) and Mahadevia (2011). of the state's own behaviour is not exceptional or residual, but rather that informality has become a mode of governance integral to the process of urban development itself (Roy, 2005;. Taking this informality as a given constituent of governance-as-practised, we focus on its implications: what does informality's presence mean for a gender-sensitive analysis of power within a slum improvement scheme? The first implication is that the presence of local political leaders, informal institutions, and improvised practices is not automatically parasitic or indicative of corruption, but may instead be important in resolving the contradictions and gaps present within formal governance processes themselves. Following Bénit-Gbaffou (2011), we should expect there to be blurred boundaries between officially devised mechanisms to ensure participatory governance, and preexisting structures of power and patronage. Furthermore, these boundaries are likely to be the points at which potentially contrasting ideas of 'proper' gender roles are revealed and contested, with ambivalent outcomes for women (cf Beall, 2005;Datta, 2012). Rather than making quick normative judgments based on whether these roles and practices conform to formal ideals, we first need to understand how administrative operations take place 'on the ground'. Accordingly, when we reflect on the outcomes of women's participation in local governance below, we do so via a contextualised understanding of the intersection between the institutional mechanisms through which governance is expected to take place, and the norms, values, and power relations which shape women's ability to act.
The second is that, if governance-as-practised involves state institutions and other actors in a process of improvisation, then we need to examine how these potentially fluid administrative operations gain legitimacy and become lasting. Lund's work on 'twilight institutions' provides a useful guide here: he sees "the making of public authority as an active and contested process of assertion, legitimization and exercise" (2006a, page 679), in which authority is continually built up (and challenged) by participating in struggles to define and enforce collectively binding decisions and rules. Authority is therefore a performative act, and one in which stability is sought through processes of 'symbolic bricolage', or the borrowing and reuse of existing symbols of power deemed to be legitimate. These borrowings might include state actors and institutions attempting to naturalise their power through appeals to ideas of tradition or locality, or conversely informal actors and institutions attempting to demonstrate their 'state-like' nature by adopting the form, language, or practices of their formal counterparts (Lund, 2006b, pages 690-693). Lund particularly highlights "the capacity to make distinctions … [as] the essence of public authority" (Lund, 2006b, page 689). The ability to define, enforce and develop legitimacy for the boundaries that differentiate between "citizen and stranger, owner and squatter, violence and punishment, acceptable and unacceptable, and so on" (2006b, page 698) is therefore key to the establishment of de facto authority. Observing these distinctions, the gendered processes through which they are produced, and their effects in excluding/including people as the legitimate subjects of government attention therefore becomes central to explaining how this slum improvement scheme unfolds, even though its context is far divorced from the African examples of Lund's study. (3) This paper therefore sees governance practices as continually improvised, and analyses them without an a priori privileging of either 'formal' or 'informal' practices, an approach that in turn has two consequences for our analysis. The first is to look more inclusively at the range of institutions, individuals, and practices involved in actually existing governance.
(3) Lund himself places caveats around his work, noting its importance for studying the state within Africa, where the boundaries between formal and informal institutions may be particularly porous, and claims to exercise legitimate public authority particularly open to contestation. Formal governance structures undoubtedly are more stable in Kerala than several of the contexts present in his collection, but we argue that his emphasis on improvisation, bricolage, and the ability to make binding decisions provides an important entry point to an understanding of governance as practised.
We therefore describe the official structures through which housing beneficiaries were to be identified, and the state's expectations of women's roles within them-an important issue as state and NGO efforts to inspire women's participatory activism have proliferated in contemporary India (Datta, 2012;Nagar and Swarr, 2005). Our empirical description of how a beneficiary list was actually developed and made operational is, however, necessarily broader. It includes not only 'officials' but also a range of elected representatives, informal leaders, fixers, and brokers, and the practices involved similarly spill out beyond the formal spaces of the city council, its ward meetings and official engagements with beneficiaries, to a range of individual actions and public performances aimed at pressing particular claims.
The second is to pay particular attention to the justifying registers through which public authority is claimed. We start here with Hansen's (2004; observation that India exhibits three parallel repertoires for making power legitimate: the law (appeals to a sublime Weberianrational state), 'civic individuality' (the assertion of effective personalised authority, often by a local 'big man'), and the community (as the embodiment of a supraindividual will).
Each is inherently open to challenge: the law through the state's own informal behaviour; individualised authority either from opposing claimants, or from the partiality of the violence on which efficacy is often based; and 'communities' because they are necessarily fictitious simplifications of social reality. Each, however, provides powerful resources for claim making through bricolage, and where we wish to push Hansen's analysis forward is by examining the place of gendered identities inherent within these symbols and their deployment.
Drawing these together, we aim to fill out the state's own sketchy vision of JNNURM project implementation, indicating the dependence of this vision on a 'supporting cast' of informal actors and the improvised scripts that guided their behaviour. Furthermore, we recognise the interplay between the actions of key protagonists and the personal circumstances that enabled them to play 'public' roles. These were necessarily shaped by their positions within the slum and their own households, and as such involved negotiating spaces of gendered power shaped both by the state's own designs and by evolving social norms (cf Das, 2011;Datta, 2012).
Making political space for women? Kudumbashree, city governance, and slum improvement Kerala has been internationally celebrated as a leader in participatory governance since the unfolding of the People's Planning Campaign in the 1990s (Heller, 2001;Thomas Isaac and Franke, 2000;Thomas Isaac and Heller, 2003;Törnquist, 2004), and it has had the most farreaching programme of decentralisation within India (Heller et al, 2007). Here, we briefly outline how decentralisation interacts with Kudumbashree, the state's poverty alleviation programme which began in 1998, to map out the official urban governance structures through which 'our' slum improvement programme is to be implemented. India's 74th Constitutional Amendment (1992) made elected municipal councils mandatory: Kerala was among the first to adopt this via the Kerala Municipality Act (1994, with amendments in 1995, 1996, 1999, and 2000), and from 2001 ward committees have been an active third tier of governance in all towns with populations over 100 000. The composition of the ward committees (table 1) is supposed to represent a broad spectrum of local 'stakeholders': they have a range of powers and functions from assisting social welfare programmes to the implementation of spatial planning (TERI, 2010). They are also required to draw up local development priorities (through public meetings or ward sabhas), and significantly the municipality is not supposed to alter development priorities set by the ward committee. Kerala's capital, Trivandrum, which has a population of around 960 000, reflects this within its governance structure (figure 1). Its municipal corporation consists of 100 wards, of which a third are reserved on a rotational basis for women: ward councillors are directly elected every five years, and they elect the city mayor who is supported in running the city through a series of seven standing committees (finance, development, welfare, health and education, public works, town planning and heritage, appeals). This structure has been filled out and supported by Kudumbashree, which brings federated groups of 'below poverty line' women into city governance. Neighbourhood groups (NHGs) of fifteen to forty women form Kudumbashree's lowest tier, cemented through activities that include collective savings and credit: an elected representative of each NHG participates in an Area Development Society (ADS) at the ward level, and each ADS in turn elects a chair who participates within a municipal-level Community Development Society (CDS), which is a formally registered NGO.
Significant responsibilities within urban governance have been handed over to Kudumbashree. At the state level, the Kudumbashree office has been declared the State Urban Development Agency. Since 1998 CDSs have also replaced the city-level Urban Poverty Alleviation Cells previously active within Kerala's municipalities, and hence have responsibility for implementing various central government urban poverty alleviation programmes (4) (Government of Kerala, 2010). Neighbourhood groups, Kudumbashree's lowest tier, are also significant actors within the ward committees. This reflects wider changes across Kerala to use Kudumbashree both to provide an integrated approach to poverty alleviation, and to deliberately create 'invited spaces' for poor women's engagement with planning processes.
Kerala has therefore deliberately embedded women within a broadened 'institutional surface area' of urban governance, but these 'invited' participatory spaces inevitably carry their own cultural and political baggage. Kudumbashree was established by, and remains closely associated with, Kerala's Communist-led Left Democratic Front (LDF), meaning that this was seen as an instrumental attempt by the LDF to use participatory governance to dominate its political rivals. In response to this perceived bias, formal elections have been held since 2008 for its higher level posts (the ADS and CDS tiers), (5) but in so doing Kudumbashree women (4) These include the Swarana Jayanthi Shahari Rozgar Yojana, an urban self-employment programme, the National Slum Development Programme, and Valmiki Ambedkar Awas Yojana, a housingprovision programme for slum dwellers. (5) Although elections in theory give all parties equal opportunities to ensure that their women supporters are actively engaged within Kudumbashree's structures, the CPI(M) (Communist Party of India, Marxist) still benefits from its association with the programme's foundation. In 2006 the Congress Party launched Janashree, its own federated network of savings and credit groups, to challenge the CPI(M) in this regard-but Kudumbashree remains the only network that has an official role within planning and governance.  have been drawn into the fiercely competitive arena of local party politics, with mixed results (see Williams and Thampi, 2013;Williams et al, 2011). Questions have also been raised about the ideals of femininity promoted within Kudumbashree, for the increased 'voice' it provides to women is accompanied by their expected adherence to conservative norms of upholding family values and giving service to their community (Megarry, 2014). As with Kerala's reservation of seats for women within local governance institutions, many women's participation in public life has been structured within a particularly feminised and 'apolitical' role, whereby they aspire to be rule-abiding and sensitive distributors of welfare (Devika and Thampi 2012;Devika et al, 2008). Those able to transcend this role are often holders of considerable social capital, in Bordieu's sense of the term, through professional status, or the support and acculturation that came from being from a 'party family'. Conversely, poorer women, and particularly those from low-caste backgrounds, often struggle both financially and socially to conform to the performances of respectability associated with being a 'good' Kudumbashree woman. Within slums, women additionally have to act within an urban political sphere rendered complex through both the absence of a stable community to whom they can be the 'generous giver' of welfare benefits, and the presence of high-stakes battles over urban development. Kudumbashree therefore provides women in slums with institutional pathways to power (Spary, 2007), but utilising these requires knowledge and strategic negotiation of an institutional and political environment that is far more complex than that facing their rural council (panchayath) counterparts (Devika and Thampi, 2012). Given this wider linking of governance, welfare, and women's participation, it is perhaps no surprise that, at the launch of the JNNURM in 2006, Kudumbashree was declared the State Level Nodal Agency for the implementation of the Basic Services to the Urban Poor (BSUP) for Kochi and Trivandrum, Kerala's two cities that have their urban redevelopment supported through the programme. Nationally, BSUP projects are approved through the Ministry of Urban Employment and Poverty Alleviation, and central government grants form a significant proportion of their funding. (6) Within Trivandrum's City Development Plan, slum improvement comprised almost 60% of the total INR 5.77 billion (US$95 million) planned expenditure within the BSUP between 2006 and 2012 (Government of Kerala, 2010), beginning with physical upgrade of four pilot projects, of which Kulamnagar was the largest. A locally respected nonprofit organisation, SUSTCO, was commissioned to deliver the Kulamnagar project: it proposed demolishing existing housing, replacing it with walk-up apartments of a design inspired by the celebrated architect Laurie Baker. This innovative plan, which included provision of community facilities (a community hall, marketplace, and workspaces for microenterprises), gave this project a high profile both within Kerala and within the JNNURM nationally. Understandably, however, it was its provision of housingformal, serviced three-roomed apartments that become saleable private property after seven years (7) -that was foremost in Kulamnagar residents' minds.
Official project housing allocation mechanisms reflect both JNNURM stipulations, and Kudumbashree's close integration with city governance structures shown in figure 1. Programme implementation is overseen by the elected councillors of Trivandrum's Welfare Standing Committee, but the allocation of housing is devolved to the local level. The ward councillor is responsible for overseeing this process, but a locally elected Cluster Development Committee should produce a beneficiary list. The local Kudumbashree (6) For Trivandrum this is 80%, the remaining balance being provided by the state, city, and, where relevant, beneficiary contributions to the projects (Mahadevia, 2006). (7) Allottees were to make a contribution to the cost of their apartment at the point of transfer and occupation-INR 25,000 (US$420). Although substantial, this represented only a small fraction of the building costs, let alone the potential market value, of the apartments.
Area Development Society has an ex officio place within this committee, giving local women a key role in identifying legitimate claimants. Beneficiary lists were to be ratified by the ward committee before being passed on to the city council for final approval. With Kudumbashree neighbourhood groups being strongly represented within the ward committee, official governance structures were again intended to ensure that local women, mediated via Kudumabashree, would have significant input into this process. Eligibility criteria for the housing were, ostensibly, clear within Trivandrum's JNNURM planning process: housing should be given to adults who could demonstrate that they were local residents, below the poverty line, not in possession of private property elsewhere, and were married with family members dependent upon them, with priority of allotment given to those most in need. Local implementation of these instructions was, however, sufficiently difficult to bring to the surface a range of contests over authority.

Contesting authority in a central Trivandrum slum
Kulamnagar itself is one of Trivandrum's larger slums, located close to the city's main market in an abandoned sacred grove (kaavu) that had become a waste dump. Although prone to seasonal flooding, from the mid-20th century a mix of Muslim and low-caste families settled there, and early-established residents rented out land and/or shacks to those who came later (cf Das, 2011;Datta, 2012, page 88). From the 1960s its official status was that of 'government land' marked for the development of sewerage infrastructure, but its central location meant that an informal property market quickly emerged, with shacks changing hands for considerable sums of money. (8) Although the majority of these informal transfers were honoured, from the mid-1970s any resulting conflicts were being resolved by Comrade Madhavan, a local Communist leader who represented workers in the neighbouring market. He became an important figure within the slum, and through his intervention residents were issued with ration cards, certificates of possession, (9) and basic infrastructure in the 1980s. Significantly, this unofficial role played by a Communist leader produced state recognition of the slum, rather than its formalisation: these gains seem to have been sufficient to end demands from within the settlement to secure full title deeds. Kulamnagar has also long been the target of government 'improvement' measures: in the early 1960s (an era when sanitation workers were unionised and launched effective strikes), around 100 of its Scheduled Caste (10) households were relocated to a government-built resettlement colony, and in 1981 a set of government flats were built on the site in a first phase of reconstruction and formalisation. Alongside this developmental role, government has also long been felt as a repressive force within the slum, through attempts to evict slum residents, and also through police raids and extortion based around the illegal arrack (alcohol) production and drug trade for which the area had become notorious by the 1970s.
(8) Our oral history work shows how prices of huts have appreciated: one Kulamnagar-born resident remembered her grandmother purchasing her shack at INR 150 (US$2.45) in the 1950s-the same shack was sold for INR 33 000 (US$1045) in 2000, and then repurchased for INR 100 000 (US$1635) in 2004. Another household reported spending INR 250 000 (US$4090) on replacing their shack with a two-storey brick-built structure. All exchanges and improvements occurred in the absence of formal title deeds. (9) Certificates of possession are not property titles, for Kulamnagar officially remains state-owned wasteland, but do confer recognition of residence by Trivandrum City Council. When slum houses are traded, for the new owner to be recognised, the original holder needs to register her or his voluntary transfer of the certificate of possession. In practice, this often was not undertaken, meaning that there was not a one-to-one correspondence between 'owners' of slum housing and those recognised as slum residents by the city. (10) Formerly 'untouchable' communities at the bottom of the Hindu caste hierarchy, which the state has constitutional responsibilities to support.
Perhaps unsurprisingly, Kulamnagar's population faces marginalisation from 'mainstream' society. The stigma of its slum status and notoriety is reinforced by the illicit activities (arrack and drug sales) or legal but 'polluting' work (sanitation and leather working) that have been long-standing parts of the livelihoods of many households. Both intersect with Kulamnagar's social composition: low-caste and Muslim households predominate, but Kulamnagar also has a disproportionately high number of intermarriages (Madhusoodhanan, no date) and a significant number of incomers from 'downwardly mobile' families and women fleeing domestic violence from spouses, all compounding social exclusion in a context in which family break-up and intercaste/intercommunity relationships are considered shameful. Challenges to this exclusion have emerged throughout Kulamnagar's history: some early settlers, both men and women, were active within formal political associations (particularly the Communist movement) fighting for social and economic rights, and Comrade Madhavan negotiated in a 'patron'-like manner for state benefits for the slum over the 1980s. Since then, a generation of women (many now in their forties) have spent their adult lives engaged in community activism, beginning with participation in Kerala's Total Literacy Campaign (1990), then a local anti-arrack and drug drive that gained city-wide attention in the mid-1990s, and participation in Kudumbashree groups from the late 1990s. The 'outcaste' status of slum women has given them licence to participate in this public realm since Kulamnagar's inception, albeit at high personal cost, (11) but has also distanced them from precisely the forms of 'gentle persuasion' that Devika and Thampi (2012) note are expected of 'respectable' female local politicians or Kudumbashree leaders elsewhere in contemporary Kerala.
The JNNURM upgrade programme came to Kulamnagar in part because of this history of 'notoriety' combined with community mobilisation for social reform. SUSTCO saw the potential of the site, and in 2006 was invited by Trivandrum City Corporation to speedily produce a Detailed Project Report for its redevelopment. Their entry point into the community was Kartika, a local woman who had built up a reputation as a social activist through her involvement in the anti-arrack drive of the 1990s, followed by a decade of leadership experience within Kudumashree. She helped to bring residents together in a series of public meetings where details of the upgrade programme could be presented and discussed. Part of SUSTCO's task in drafting the Detailed Project Report was to list the households eligible for rehousing, a job that was given to local Kudumbashree women. Again, Kartika, as the elected chairperson of the Kudumbashree ADS that served Kulamnagar's ward, was central to this, and she and her local coworkers after just two days of house-to-house surveys produced a list of 430 families owning houses. SUSTCO's executive director instructed her to expand this to include also those who were renting property within Kulamnagar and, following quick amendments, a beneficiary list of 560 households that met the JNNURM eligibility criteria was produced, passed by the ward committee and approved by the corporation council (SUSTCO director interview, 16 July 2013). Residents were given fifteen days to contest the list, which removed sixty-six names and replaced these with others, but as at this point in time many either did not believe that the project would materialise, or did not see the importance of ensuring their inclusion, this revised list has also been challenged subsequently. Significantly, however, the figure of 560 households as a total population for whom housing should be provided has not been adjusted: although this was based around a quick and inevitably approximate survey, it became a fixed delivery target through the JNNURM budgeting process.
(11) See Devika (2013) for a fuller history: although most Kulamnagar women have no 'respectability' to lose through being strident, coarse, or even aggressive in their public demeanour, oral histories repeatedly indicated that their 'freedom' to be politically active independently of their families was bought at the expense of exposure to domestic violence.
The local MLA (12) (a Communist Party member, and ex-Mayor of the City) provided the initial push for the project's implementation, forming an informal cross-party committee of local leaders to guide the redevelopment, and ensuring that Kerala's Chief Minister was present at its official launch in 2008. With Kartika as a competent ADS chairperson also actively involved in local implementation, the formally required mechanism of a Cluster Development Committee (CDC) initially was either not constituted or not activated. Perhaps unsurprisingly, however, as the first apartments were built, contests over their allotment heated up. The SUSTCO plan had envisaged that demolition and rebuild would occur in four phases, and there was anger that within the first two, the councillor who represented the ward from 2005 to 2010 (whilst it was a female-reserved seat) had unfairly favoured her own political supporters in the allocation of flats. Kartika, who herself was allocated a flat in phase 2 (2009-11) of the rebuild, was seen as either complicit in this process, or unable to resist the councillor's interference. Even though Kartika had actively encouraged community involvement in phase 2 by demolishing her own hut, (13) when she took possession of her new flat, she was threatened with attack from Jadeeda, a Kulamnagar resident with connections to the SDPI, an Islamic political party. The disruption at this point brought negative media attention to the slum, which highlighted its violent and 'ungovernable' nature, and subsequently a CDC was formally elected from local residents with a range of different political party backgrounds.
What we see up to this point in the project's history is a far more complex picture of actually existing governance than the official picture presented in figure 1. First, there was ongoing and unresolved tension over both who was included in the original eligibility list of 560 households, and in the particular allotment lists drawn up for all four phases of programme implementation. JNNURM criteria aimed to demarcate eligibility within the programme as a whole, but official records and ground realities did not always match. (14) Criteria for allocation of flats within each particular phase were not specified beyond putting the most deserving first: households directly affected by the demolition were included, but beyond this the CDC had to determine an appropriate priority. Residents, some of whom had taken little notice of the original 2006 household survey, now actively sought to secure their position within both eligibility and allocation lists. Second, the official practices aiming to control this process were improvised: significantly, the city corporation activated the CDC only when it became clear that leaving allocation to Kartika and the local Kudumbashree ADS was leading to conflict. Finally, there was a complex and multilayered party-political dimension to these conflicts, in which claims to legitimately represent the interests of Kulamnagar as a community were contested. For political leaders, from local activists such as Kartika through to the city mayor, the project offered public recognition and a demonstration of their authority. At the same time, claims of 'political bias' in flat allocation were a potential source of challenge that could both question their personal legitimacy, and stall the project's successful implementation.
(12) Member of the Legislative Assembly-the state-level parliament of Kerala. (13) There was a logic to Kartika's action: when residents demolish their own houses, valuable materials-from door frames to electricity cabling-are recovered for sale or reuse, and these would be lost if their properties were bulldozed. (14) Residence could usually be demonstrated through existing forms of government recognition (possession certificates and ration cards linking a household to that address), although as noted in footnote (9), possession certificates and actual residence did not necessarily match. Property ownership outside Kulamnagar might be 'known' locally, but proof was more difficult, and opened up a route through which inclusion/exclusion could be contested. Finally, proof of heading a family pitted state ideals (a married husband, his wife, and their heirs) against Kulamnagar's realities of separation, cohabitation, and filial abandonment of aged mothers: interpretation here was therefore particularly contentious, with natural justice often clashing with patriarchally informed 'official' status.
Our own fieldwork allows us to take up the story of the allotment of the phase-3 flats, from early 2013. By this stage, the ward councillor who had been accused of corruption in phases 1 and 2 had been replaced by Mr Pillai, who was elected in 2010 as a Congress Party member. High-caste, and predominantly focused on the middle-class constituent members of the rest of his ward beyond Kulamnagar, he nevertheless sought to exert some control over the project's implementation. Other protagonists were, however, unchanged: Kartika remained the local Kudumbashree leader and an active CDC member, Jadeeda was still trying to use her political connections to the SDPI to gain flats for her extended family, and Comrade Madhavan sought to maintain his patron-like control of the settlement, even though it had not been officially part of his own ward for over a decade. Our research began soon after the Kartika and her fellow CDC members had worked hard, using their local knowledge, to produce an allotment list of the most deserving 180 households who were to be phase-3 flat recipients. Through our first-hand observations and interviews with participants, we review the struggles over this contentious document, how the protagonists within them sought legitimacy for their actions, and the role of gendered identities within these practices.

The phase-3 inauguration meeting (February 2013)
We begin with the phase-3 inauguration meeting, held in a community hall immediately outside Kulamnagar, to which all families identified as flat allottees by the CDC were invited. These beneficiaries should have been a clearly demarcated group, as all had been issued with an allotment certificate by the CDC over the previous weeks, but policing attendance was a significant task in itself. Kartika and other members of the CDC attempted to exclude those entering the hall without these certificates, through strongly worded threats, and directing the anger of those present against any individuals clearly identifiable as 'outsiders'. Forty-five minutes after the meeting was due to begin, and with the audience somewhat settled, the dignitaries arrived and took their places on the dais. The Vice-Chair of Trivandrum Municipal Council and the chairs of the city's Health, Finance, and Welfare Standing Committees chair took turns to address the crowd (the ward councillor was also present, but silent): after berating those present for their unruly conduct, they each presented the project in terms of the city's generosity in enabling them to uplift their living conditions. The project's tight time-frame for implementation was repeatedly mentioned: with funds lapsing in 2014, the city expected that the community would cooperate fully in its smooth implementation. This need for compliance was presented alongside some fairly direct threats, aimed not just at maintaining order within the meeting itself ("If you fight and yell, we will go away and I'll report to the mayor that this project should be closed"), but also at removing false claimants for flats ("We have our own ways of finding out … . We will surely make inquiries to find out if persons in the list hold land anywhere else in the city and weed them out"). Pointed mention was also made that there would be high-tech biometric recording of beneficiaries when demolition commenced to eliminate fraud. A space was provided for residents to express their views, at which point a woman who had not gained flats for all her adult daughters rushed to the microphone to demand their names be placed on this beneficiary list: her claims were hotly contested by Kartika, and she was expelled from the meeting ("How did you dare enter the hall-you have no allotment slip!"). After order was restored, the Finance Committee chairman assured those present that all valid claims would be properly examined, and a date to commence demolition work was set for 10 March. Significantly, there was only the vaguest mention of providing temporary housing for those displaced during the rebuild, and no firm suggestions of how this was to be implemented. Although this was pressing concern for many of the households present, they preferred to individually lobby Kartika for help after the meeting, rather than demand this of the council representatives present.
The city councillors presented the project as operating within the law and formal processes to the benefit of the community as a whole, but it is interesting that this 'official' picture was disrupted in various ways. The failure to provide temporary housing was a transgression of the state's own rules that risked stalling the whole project, (15) but more generally this (relatively) smooth display of the state's power was itself reliant upon other forms of authority. Kartika and other members of the CDC were clearly important gatekeepers here: they 'policed' the venue, and enforced their views of who should be regarded as valid beneficiaries, which were justified in turn through appeals to a moral economy which valued 'deserving residents' over 'speculative' claimants. (16) By allowing her verbal expulsion of a 'false' claimant to stand unchallenged within the meeting, leading members of the city council were implicitly confirming her authority to set such boundaries. In a situation of intense competition over housing, far from perfect official records, and vague eligibility criteria, they were happy to subcontract the contentious work of prioritising claims to others.

Conflict within Kulamnagar (March 2013)
Kartika's power exerted within the public meeting was, however, far from unopposed. She had to contend with the continued presence of the ward's former councillor, Comrade Madhavan, but also with Jadeeda and her supporters who had challenged her allocation of flats within phase 2. Madhavan's undermining of Kartika was subtle: he held an additional public meeting, allegedly at the personal behest of the city mayor, to ensure that any households threatened by demolition that contained severely disabled members were immediately rehoused. Although their needs had been previously identified by Kartika, he managed to present himself publicly as someone who could 'resolve' this issue speedily and without conflict. He contrasted his own calm and unquestioned authority with Kartika's strenuous efforts to maintain control, claiming that Kulamnagar's community "can't be held in place by women, it won't fit into their palms" (interview, 13 March 2013). The challenge of Jadeeda and her supporters was far more direct: they established the 'Kulamnagar Colony Action Committee', and drew on their linkages to the SDPI to disrupt progress. One action here was to hold a rally outside the state Legislative Assembly (21 March 2013) to argue for transparency in the allocation process, and to publicly and personally accuse Kartika of corruption in setting the beneficiary list. Significantly, however, this rally was not able to draw wide support from within Kulamnagar itself, and the majority of those present were male SDPI activists.
When the rally proved relatively unsuccessful, supporters of the group formed a mob that threatened the chair of the CDC. They gathered outside his house at night armed with swords and chilli powder, (17) and things looked sufficiently grave that Kartika called both the police and Comrade Madhavan for support. According to Kartika's account, it was Madhavan who arrived first: he immediately took charge, ordering all external SDPI activists to leave the slum, and the others to drop their weapons and disperse. When the police arrived, Jadeeda (who had been present within the mob) attempted to get them to hear her claims of Kartika's corruption, but she was publicly rebuked and told to shut up by the chief officer present.
Again, we see the mixing of 'official' and highly informal modes of claiming authority here. In Lund's terms, the rival group borrowed the 'state-like' form of a committee to present (15) After the meeting one lay participant cited national law over the rehousing of 'project affected persons' and argued that he could launch a court injunction that would stall the entire process: "It would cost me just INR 5000" (US$80). (16) Official criteria for eligibility were themselves the target of speculative actions by residents: not only was there an influx of returnees, but families also attempted to gain additional flats by quickly marrying adult children, so that they fitted the state definition of a 'household'. (17) Used, in this instance, as a homemade version of pepper spray -this was carried by women within the mob.
itself as a legitimate claimant representing Kulamnagar as a community, and upholding ideals of good governance. At the same time, its participants had no qualms about resorting to violent means to intimidate the CDC chairman and challenge Kartika's authority. Kartika had the linkages to counter this through her connections to the police and Madhavan as the official and unofficial providers of security, but she was unable to deal with this threat unaided. Madhavan did so: despite being an old man, his verbal command was sufficient to cause the mob to disperse, for everyone present knew the direct threat of force it contained. (18) Enacting an 'orderly' community (April 2013) Recognising that these events threatened both her personal authority, and potentially the entire project's progress, Kartika began her own countermobilisation of the community. Under the banner of the "Phase Three People's Committee", she launched a pro-redevelopment demonstration, marching the 6 km from Kulamnagar to the city corporation office, where a dharna (sit-in protest) was held. This was a carefully managed event: Kartika, a Communist supporter herself, had ensured that Welfare Party activists from Kulamnagar headed up the procession, clearly marking this demonstration out as a cross-party event. There were also over 300 people marching, with middle-aged women the bulk of those present and centrestage in the shouting of slogans: some male youths from the slum were also present, but hung around the edges of the demonstration somewhat awkwardly. Kartika also ensured that everyone present was 'polite' at all stages, to differentiate them from the 'vulgar' tactics of Jadeeda and her Kulamnagar Colony Action Committee. This was echoed in her carefully worded address to the crowd, in which she stressed that this protest was not against the corporation council or the mayor, but was rather to assure them that the project had the full backing of the community, and was only being opposed by a wily minority supported by 'outsiders'. Eventually, the demonstrators' leaders were granted an audience with the mayor, and received assurances that the demolitions would go ahead on schedule in mid-April.
Through this demonstration, Kartika and her supporters enacted a performance of community harmony that had wider political legitimacy in part because of the skills and contacts generated through her history of activist work. These were seen first in her ability to mobilise such large numbers of women: Kudumbashree members provided the core of a cross-party crowd that which could credibly claim to be the symbolic 'heart' of Kulamnagar's community, a convenient match between the state's rhetoric on participatory citizenship and Kartika's day-to-day leadership of these women as their ADS chairperson. They were also seen in her capacity to stage a demonstration that operated within an acceptable vocabulary of protest and, in contrast to her rough handling of women at the inception meeting, to engage politely and articulately with the mayor. These skills helped to explain the trust city officials-from the police through to the JNNURM unit-placed in her to authentically represent what was happening within Kulamnagar, but her 'success' in ensuring the project's speedy implementation needs to be balanced against other aspects of the project, such as uncertainty over transitional housing, that remained unaddressed.

Predemolition recording begins (May 2013)
Two months after the intended start date, recording of the people about to be displaced by demolition began. Rather than a 'biometric' process, a small team of officials with digital cameras were deployed to photograph allotted beneficiaries next to their current houses. The officials' lack of local knowledge meant that they needed to have houses and families (18) In a direct example of power of a 'boss' (Price and Ruud, 2010) or dada [literally 'big brother' (Hansen, 2004)] we were told that a single missed call made from Madhavan's mobile phone "could bring 3000 supporters running here within ten minutes", a claim made to us directly by Madhavan himself (interview 13 March 2013), but repeated by many other Kulamnagar residents throughout our fieldwork.
identified by Kudumbashree women, who again were important in 'adjusting' the scheme's official eligibility and allocation criteria to fit local ideas of justice. In one instance, they intervened to exclude a son who had returned to the slum after a long absence from being recorded as his mother's heir, as they feared that this official recognition would embolden him to turn her out of the flat as soon as she was given it. The recording almost descended into farce when one of the officials' memory cards ran out of space, and the camera on the smartphone of one our research team had to be used to finish the process.
These somewhat haphazard operational details were themselves a far cry from the picture of a sublime, authoritative state presented in the inauguration meeting, but behind them lay more fundamental questions about the security of the whole allocation process. In the four weeks between Kartika's demonstration and the beginning of recording, Jadeeda and her supporters once again tried to disrupt the entire project. They attempted to intimidate those on the beneficiary list not to sign the demolition consent forms necessary for the rebuild to begin, and also provoked a further direct attack on Kartika, when an assailant armed with a meat cleaver broke into her house and threatened her. During this period Comrade Madhavan had also taken up residence in the slum "to ensure that there was no violence", and encouraged those not on the beneficiary list to register themselves within an appeal process, an action which created a parallel list of fifty-one households. The city's response to this ongoing resistance was a fudged compromise: rather than simply sticking to the original list (which now had the formal approval of the city council), the additional fifty-one households were recorded as 'potential beneficiaries'. This compromise moved construction forwards, but also set in train raised expectations and another category of quasi-recognised claims, further complicating the settlement's future redevelopment. Even as the latest round of demolition and rebuilding of Kulamnagar began, questions of who had an official claim to a flat, and who ultimately had the authority to make binding decisions over their allocation, were thus being renegotiated and deferred, rather than resolved.

Making space for women, gendering urban governance
These continuing battles indicate that official practices and procedures were insufficient in and of themselves to deliver this slum improvement project. This in part relates to the project's 'top-down' elements: its timeline was driven by national funding deadlines, and SUSTCO had produced an 'inspirational' (but inflexible) design that would deliver only 560 housing units, setting some hard constraints around which implementation had to operate. These constraints were difficult to reconcile with the fluid realities of informal life in Kulamnagar, with its dynamic household composition and migration, and further complicated by both the 'messiness' of formal records of residence and the vagueness of eligibility and allocation criteria. Together these meant that some degree of localised discretionary power was a necessary component of the project, but in a circumscribed and particularly contentious role: transforming an externally unknowable population of potentially eligible beneficiaries into a stable allotment list of households.
The state's formal response to this problem was to give Kudumbashree women a role in beneficiary identification through the CDC. In practice, this was initially left to the local Kudumbashree ADS, and particularly Kartika as the project's community organiser. The attractiveness of Kartika to the state was clear: her history of activism and intimate knowledge of Kulamnagar's inhabitants allowed her to become a 'bridge' between the project and the slum, and her elected position as the ADS chairperson gave her some semblance of democratic credibility. This in turn ensured a degree of gender-sensitive interpretation of eligibility criteria, with Kartika and her compatriots able to marginalise male heirs who had abandoned their mothers, or promote the claims of destitute women. This exemplifies Kabeer's (1999; definition of empowerment, the expression of choice to challenge patriarchal norms, delivering significant change in the interests of gender justice for those whose position within the allocation process was directly affected. However, it is equally important to note that this expression of agency did not change the project's broader parameters, and carried a high personal cost, as the repeated threats of violence towards Kartika and the CDC chairman showed. Looking beyond these immediate effects of 'improvised' governance practices, we turn to the implications of this case for the theorisation of gender within political authority. First, in order to explain how the main protagonists in this story to came to play 'public' roles, it is necessary to understand Kulamnagar's history, and the interplay of 'slum' identity and female-centred social activism within this. Women there had benefitted from external attempts to organise and empower them (beginning with the Total Literacy Campaign in 1990) and were also transgressing the norms of 'polite' Keralan society, through economic necessity, family break-up, or simply their residence within the slum. These conditions had drawn a generation of women into active engagement in Kudumbashree and local politics, who were then able to take up the opportunity to participate-individually and collectivelyin the struggles for housing allocation that the project opened up. But it should be noted that this was not a linear process of emancipation within what remains a highly patriarchal society: being active within the public sphere often went hand-in-hand with suffering genderbased violence within the home (cf Megarry, 2014), and gender-based slurs and questions of sexual morality were recurrent themes deemed locally to be 'legitimate' grounds on which to criticise women like Kartika.
The broader point here is that to anticipate the impacts of development interventions promoting community participation-such as this JNNURM project-we need to understand actually existing governance arrangements. The landscape of individuals and institutions active within struggles to claim authority within (or over) the community is shaped by economic opportunities and constraints, social norms enforcing abjection or inclusion, and preexisting forms of solidarity or division within the community. This in turn requires careful spatial and historical analysis, for women's engagement with 'invited spaces' of participation does not occur within a vacuum, but emerges through their experiences of the many layers which constitute this landscape (cf Das, 2011;Datta, 2012). Furthermore, gender inflects both the performance and the reception of women taking up public roles (Spary, 2007), an observation we develop further here by thinking through the role gender plays within the legitimating registers of the community, civic individualism, and the law (Hansen, 2004) that help to construct public authority.
Within Kerala, Kudumbashree provides important practical opportunities for women to enact ideas of community, albeit ones which are shot through with gendered assumptions about the 'public service' they should embody. Kartika's protest march shows the potential power of playing with, or playing up to, state-backed ideas of women being the 'heart of the community', but also the alternative constructions being placed on Kulamnagar that had to be supressed. The party politicisation of much of Kerala's public life meant that cross-party support had to be a central part of this performance, in order to avoid being portrayed as partisan or self-interested. Slum identity also returns here, for the threat of the whole of Kulamnagar being labelled by the city council as 'unruly' (and thus unworthy of support) hung over the project.
This idea of community is a substantial distance away from Hansen's focus on the processes reifying differences between caste and religious groups, and the riots and pogroms that can ensue where the interests or honour of 'the community' defined according to these differences are deemed to have been harmed (see also Brass, 1997). Such violence certainly has lasting effects on slum populations and their relationships with the state (Das, 2004), but we would like to argue that the more nuanced and everyday performances described here are equally important. Governmental categories (Kulamnagar as a territorially bounded 'slum cluster', or the 'poor women' who populate Kerala's Kudumbashree programme) often cross-cut other forms of collective identity such as caste, and it is through the need to negotiate these intersections that new forms of political community are being improvised, in ways reminiscent of Chatterjee's (2004; work on political society (see also Gurudvathy, 2011). We would not wish to romanticise the transformative effect of this, for Kulamnagar remains highly marginalised within Trivandrum and as an 'object' for state intervention, but successful performances of an orderly community remained important in bolstering claims to public authority in project implementation.
Turning to 'civic individuality', Kartika provides an interesting case of how a marginalised woman can attempt to assert effective, personalised authority. She certainly sought out leadership, and was effective in using the 'rough talk' of the slum to challenge opponents in the lanes of Kulamnagar, or to silence them in public meetings, countering mainstream Keralan ideals of the polite, apolitical Kudumbashree woman as she did so. However, this vociferous enforcement of her will was compared unfavourably by many slum residents with Comrade Madhavan's quiet powerfulness, backed of course by his command of force. Kartika's role arguably represents a particularly gendered path towards curtailed 'civic individualism', in which links to officialdom were both necessary to build her authority, and yet insufficient to secure this. Her connections to the police, the city council as well as its Kudumbashree offices were not only instrumentally useful for the project, but also provided her with the 'official' status of project community organiser, and a degree of broader legitimacy through the trust these city authorities placed in her. But this position also had its limits, and was insufficient to contain dissent and make decisions binding, especially when faced with violent disruption: to use Comrade Madhavan's idiom, if her connections put the community "in her palms", she was unable to hold it there.
The wider implications of this case are that the literature on leadership in South Asia needs to engage with women's growing presence within local governance. Earlier scholarship on political 'fixers' in South Asia (Manor, 2000;Ram Reddy and Haragopal, 1985) simply assumed men's centrality to these roles: within other work, masculinity (and machismo) emerge as important sources of political bosses' power (Brass, 1997;Hansen, 2004), but detailed studies of how women exercise authority remain rare (although see Spary, 2007). As Price and Ruud (2010) note, individual leaders may build legitimacy through prowess/ dominance (within which violence is particularly important), but also through generosity/ patronage and/or 'insider knowledge' of official governance structures. They usefully highlight the ambivalent but also productive role of informal leadership, but what they do not do is to investigate the degree to which each of these sources legitimacy might be more or less open to use by women. The continued presence of local bosses-and their ability to achieve efficacy through control of violence-may necessarily place limits around the power afforded to others within participatory governance structures.
This in turn links directly to the law as a means through which claims to legitimate authority are made. The 'informality' of state's own actions was important here, first in its shifting position over Kudumbashree's importance in beneficiary selection vis-à-vis the CDC, and then by not backing the decisions Kartika made within the CDC in phase 3. It was this failure to fully support its own local body that left the space for alternative voicesand more traditionally masculinist 'boss'-based forms of authority-to undercut the links to 'officialdom' that were key to Kartika's credibility within the slum. The more generalisable point here is that, when women's authority is in part dependent upon their official status within state institutions, the improvised nature of actually existing governance arrangements itself can undermine their leadership. This improvisation also had its effects on individual beneficiaries, through the failure to close down the lists of counterclaimants for housing. The risk here is that procedural uncertainty opens up the space for rival mappings and surveys of the 'target population', in what becomes "a parody of Foucault's model of surveillance because what it generates is the emergence of new forms of brokerage that drain people of their meager resources" (Das, 2011, page 331). Rather than sublime Weberian rationality, we have the operation of Gupta's (2012) far more partial and prosaic 'red tape': a state that lacks institutional and informational unity, and as such is inevitably arbitrary in terms of the effects of its interventions into the lives of the poor.
The wider practical lessons for experiments in empowering women within urban local governance are that we need to pay particular attention to the gendered ways in which power is performed. This certainly means moving beyond simplistic ideas that women can authentically voice 'the interests of the community' uncontested. They can use a combination of individual standing and institutional linkages to fill out incomplete elements of state-led development visions, making difficult decisions and articulating a sense of justice informed by local norms/values as they do so, but expressing choice to challenge patriarchal values also carries significant personal risks. Although it is not possible for the state to erase at a stroke deeply embedded unequal gender norms, it must pay full attention to their likely effects in shaping existing patterns of authority, power, and legitimacy. This means not merely 'making space' for women within the official structures of urban development projects, but also having the capacity and will to shore up the autonomy granted to them within the far more complex landscape of governance-as-practised.

Methodological note
The materials we draw on here are a series of oral histories of Kulamnagar residents (see Devika, 2013), in-depth interviews with all key protagonists, and detailed ethnographic observation by the field team, led by J Devika. Fieldwork began in January 2013 (and is still ongoing at time of writing), allowing direct observation of phase 3 of the upgrade programme as it unfolded. All interviews were recorded and transcribed in Malayalam, and analysed in English translation: J Devika's fieldnotes on the team's extensive interactions within Kulamnagar and related government offices were written in English.