From informal to formal: understanding gendered energy transitions through filmmaking in Cape Town and Mumbai

Abstract Provision of affordable housing and energy access is an urgent challenge in India and South Africa. This study adopts a participatory filmmaking approach to understand women’s agency in low-income domestic housing in Mumbai and Cape Town. Through their films, the women shared insights of how the transitional housing environment has impacted their everyday routines and how they are utilising the access to energy. By employing creative, participatory filmmaking practice within the female community, the research explored an alternative approach to understand energy demand. The cycles of participatory filmmaking were followed by workshops, interviews, and acknowledgement of participation. The films revealed that access to formal housing and electricity grid have led to improvements in the women’s welfare and wellbeing, showing the importance of privacy and the development of personal leisure activities. Yet echoing Murray Li, there is nothing ‘inevitable’ about women’s energy transitions and energy access does not necessarily reduce the women’s time spent on household chores: in Mumbai, women still cook 3–4 times per day, every cooking moment taking 1–2 hours, amounting to women putting in 56 hours of (unpaid) work per week. The study advocates the use of film as a method in energy studies: the films were able to capture the lived experience in these low-income groups, heterogeneity of the material arrangements, the pride and skills in cooking. The films enabled the participants to see their ‘invisible’, unpaid work and to express their aspirations giving us a glimpse of what ‘empowerment’ from the women’s perspective would look like.


Introduction
Provision of affordable housing is an urgent challenge in rapidly urbanising countries like India and South Africa.In Mumbai, 40% of households live in slums.Slum rehabilitation housing is a key concern of Indian national and local policy makers who are currently committed to building 20 million houses by 2020, of which two-thirds of the stock are yet to be built under the government of India's 'Housing for All by 2020' agenda and Mumbai alone is responsible for building 650,000 affordable housing units (Bardhan et al. 2015).In South Africa, housing is seen as a materialisation of citizenship rights.Post-apartheid, the South African government embarked on a vast national housing subsidy scheme which has provided up to 4 million housing opportunities, including so called 'RDP' houses, Breaking New Ground (BNG) policy that provides housing specifically aimed at unemployed/low/unstable income people, whilst also functioning as a social process to guide people onto a path of housing security.
Transitional and slum rehabilitation housing come with an access to electricity grid.This should benefit women in particular.Women are key agents in domestic energy transitions and women's traditional role at home leads to them having far more influence than men on domestic energy consumption, especially in the Global South (Khalid and Sunikka-Blank 2017).Studies on the relationship between electrification and women's work opportunities and empowerment indicate positive results.In Guatemala, for example, women in electrified homes were reported to spend 34% less time in cooking compared to homes without electricity, and in Nicaragua, in rural households with electricity women are 23% more likely to work outside the home (Grogan and Sadanand 2013).However, as pointed by Winther et al. (2017), only few studies have been able to identify causality between electricity and changes in gender practices.Qualitative studies on women's empowerment through electrification offer in-depth knowledge but are based on single studies whereas statistical studies that have looked at the correlation between access to electricity grid and welfare indicators have only been able to provide very limited explanations.Winther et al. (2017) argue that instead of macro level indicators, we should make more household specific observations: whether electricity is used in kitchens or are lamps used when children are studying.
In mapping of feminist geographies in the global context, Blidon and Zaragocin (2019) argue that there is inequality in knowledge construction across geographies and they identify a dominance of Anglo-centric feminist geography within this knowledge production.Also, the feminist analysis of domestic technology has been Western focused (see e.g.Cockburn and Fürst-Dilić 1994;Cowan 1989; Gram-Hanssen and Darby 2018; Mochlenber and Gram-Hanssen 2020; Shove and Southerton 2000;Pink 2004).
The study focuses on Global South contexts: India and South Africa.The research accepts postcolonial criticism that questions the ideas of modernisation and development as a linear process and criticises seeing stakeholders as subjects of obligation and experts imposing on them 'the will to improve' (Murray Li 2007).In the field of energy studies, the concept of energy justice has gained ground in recent years (Sovacool and Dworkin 2014;Sovacool 2015) but only few studies have engaged with postcolonial critique.Castan Broto et al. (2018) argue that postcolonial discourse can help to question oversimplified assumptions in energy policies and its tendency towards universalising approaches.In their study of energy transitions in Mozambique, they show how the application of mobile-based systems to incorporate households into energy access, in fact constructs 'a deserving consumer who can manage debt, rather than seeking to develop models of collective ownership potentially more sustainable and democratic' .In architectural research, decolonialization discourse has commented on the conceptualisation of 'Southern' urban practice (e.g.Bhan 2019) and participatory design (e.g.Albadra et al. 2021), but it is has not yet challenged modernist forms of low income housing or looked at it from a gendered perspective.
Castan Broto et al. (2018) call for new methods as part of postcolonial discourse and advocate delivering energy as an emancipatory project while emphasising on the concept of energy sovereignty that connects energy justice and postcolonial debates on energy (see Altieri, Funes-Monzote, and Petersen 2012).The recent paradigm shift to look at domestic energy demand as social construct (see Shove 2010Shove , 2014) ) rather than as a technical issue also calls for innovative and human-centred methods to understand energy use (Sovacool 2014).This study responds to this need by adopting participatory filmmaking as a method to understand energy use behaviour.Spatiality in cities has been investigated with film (Leigh and Kenny 1996;Svoboda 2016;Penz 2017) and the use of participatory videos (Asadullah and Muniz 2015) and self-representation in video diaries (Lenette, Cox, and Brough 2015) have been advocated in recent research projects but the application to understand domestic energy practices has been limited to a small number of studies (e.g.Pink 2021;Pink andLeder Mackley 2014, 2016).Yet film can create self-distance from which it is possible to observe daily spaces and routines at home, echoing the approach by Perec (1997) to analyse everyday life at home and to question the relationship with everyday objects that surround us.In this study, the participants used smart phones to film their daily routines for five consecutive weekdays at specific times of the day, focusing on their maintenance activities, childrearing and recreational practices.The filmmaking stages were followed by a series of interactive workshops, interviews, and acknowledgement of participation, creating a cyclic process where 'the researched become the researchers' (Baum, MacDougall, and Smith 2006).This notion of participatory action research questions the traditional production of knowledge and rejects the conventional 'subject vs. object' division (see e.g.Escobar 1984).
The paper has dual approach.First, this study contributes to the new knowledge of gendered energy transitions and how gender is an intrinsic element for energy (in)equality and that there is variability of domestic practices and energy usage within similar socio-economic classes which is generally considered otherwise in literature.Second, this study uses participatory filmmaking and a combined approach of narratives and storytelling with films to understand the energy and gender dynamics in low-income housing across Global South contexts, in South Africa and India.The research asks what happens when women in Cape Town and Mumbai move to formal housing and step on the energy ladder.What are the lived experiences of the women living in transitional and slum rehabilitation housing?What are the characteristics of the women's domestic practices and how are they shaped by material arrangements, access to energy and the use of domestic technology?
The paper is structured as follows.Section 2 presents a literature view on feminist geographies and decolonializing methodologies in energy studies.Section 3 describes the methodology and participatory filmmaking in this context.Section 4 presents the findings.Section 5 concludes.

Comparative perspectives in feminist geographies
In their mapping of feminist geographies in the global context, Blidon and Zaragocin (2019) argue that there is inequality in knowledge construction across geographies and they identify a dominance of Anglo-centric feminist geography within this knowledge production.Feminist contributions to the 'spatial turn' of the Global North (cf Foucault and Miskowiec 1986;Lefebvre 1974;Soja 1996), broadly explore how patriarchal power manifests in space.Most notable are Doreen Massey's (1992) explorations of how space constructs and articulates gender, Judith Butler's (1990) gender performativity in space, and Linda McDowell's (1997) research on gender in the workplace.Created as a resistance to dominant modes of knowledge production, this feminist movement was criticised for failing to acknowledge non-white and non-western perspectives, and for white women 'it was a shock to find themselves being regarded as part of the problem' (McDowell 1993: 308).Broadly speaking, these Global North feminisms speak to reclaiming rights within space.By contrast, in sites of historical extraction and impoverishment in the Global South, 'gender and development' is prioritised.National growth in developing contexts is the focus of development, and empowerment in the Global South is considered an economic project to 'save' women from poverty through development projects (Moeller 2018: 12).This study aims to look beyond the economic project of 'gender and development' through critically exploring space as mediated by gender, family structures, and women's empowerment across Global South contexts, in low income households in Mumbai and Cape Town.
For example, when domestic practices are described, it should be considered that gendered identities and household practices are deeply anchored in family relations, 'family-scapes' (Appadurai 1996) in the Global South.This is often missed, underestimated or misunderstood in research designs formulated from Western perspective (Wilhite 2017).In this research, extended family relations in Mumbai case meant more work for the women (accommodating and cooking for more people) but also support (the participants mentioned how during the first COVID-19 lockdown in India they were pleased to get the family's support in childcare).In Cape Town, it was observed in extended motherhood: in some cases, grandmothers, not mothers, had the primary responsibility for childrearing.
The relationship between power and space as mediated by gender, has been interrogated broadly in recent scholarship, from investigations into infrastructure, housing, and feminist geopolitical frameworks to violent geographies.Such wide-ranging entry points provide valuable insight for this study, which is situated in a domestic space that has been produced and mediated by the state.Interactions in these domestic spaces bring together assemblages of physical, digital and people infrastructures, which Datta and Ahmed (2020: 69) suggest constructs the 'intimate' relationships that exist in the home, within the family, and in the city.This gendered approach to infrastructure looks beyond urban planning, to address the embeddedness of power relations and how 'questions concerning citizenship, belonging, or civic virtue may be expressed by flicking a switch, cutting off a wire, or by installing a smarter meter' (Schnitzler 2016: 130).Further critical insights into the multi-site and multi-scalar nature of power can be gained from feminist geopolitical frameworks, which explore how these interactions negotiate the relationships between intimate and global violence (Pain and Staeheli 2014).Unlike Cieraad's (1999) suggestion that the front door of home marks a 'behavioural and emotional boundary between public and private space' , feminist geopolitics explores the embeddedness of power in the 'everyday' to contradict superficial boundaries between the 'private' home and 'public' arena of geopolitics (Brickell 2012).Jackman and Brickell (2022) analysis of the 'Drone Home' for example, engages with feminist geopolitical concerns around concerns around precarity, vulnerability and intimacy and the role of human and non-human actor's in shaping the gendered spaces, processes and relations of domesticity.Feminist explorations into violent geographies similarly explore how spatial and power relations underpin gendered violence (Brickell and Maddrell 2016).'The home' is not always a private space of freedom, but a site with its own regulations in conduct and use of space (Ahmed 1999;Hooks 1990;Massey 1992;Sibley 1995), which shapes the practices, relationships, and images of the everyday (Anderson andTolia-Kelly 2004: Ciaraad 1999;Datta 2008;Hatfield 2016).This sentiment can be extended to state housing allocation itself, which in the case of South Africa, involves a conversion to 'proper' appearances and sociabilities to be 'fit to live in a house' (Ross 2015: 102).Power as exercised by people, states or institutions, is embedded within non-material actors, and as these broad ranging feminist geographical entry-points show, is present and shapes the intimate and often gendered interactions of the everyday.As such, this study explores how a material change ('from informal to formal') can result in shifts in power, which are inextricably connected to both gendered experience and gendered outcomes.
Qualitative studies often rely on single case, but this research was set up to include two case studies, one from India and one from South Africa.Rodo-de-Zarate and Baylina (2018) suggest that one of the ways that geographers could contribute to the development of intersectionality, and feminist geographies, is by analyzing the role of place and the relevance of context.In the realm of electricity and gender, Winther et al. (2017) identify the lack of comparative case studies.By focusing on the relevance of the context and by introducing a comparative perspective from India and South Africa, this study aims to contribute to a better understanding of the place and context in women's experience of housing and energy access.

Decolonializing research methods in energy studies
This research was initially planned to be centred around in-person research, primarily based on interviewing and filming by the researchers and not the participants.The COVID-19 pandemic forced the research to move online which resulted in a radical re-evaluation of research objectives (Corbera et al. 2020), data collection and collaboration with partners, opening up opportunity to critically reflect on how we conduct research and to what ends.The remote nature of the study inadvertently facilitated the opportunity to centre participants, resulting in the decolonising of research through the mutual construction of methodologies, and outcomes that were of value for both participants and researchers.Not to be conflated with postcolonialism which is primarily concerned with the social, political, economic and historical impact of colonialism, decoloniality is an epistemological concern.It builds on postcolonial, critical and racism discourse to question the processes and production of knowledge, and how they contribute to the marginalisation of people, places and thinking, which in turn reproduces 'the norms and privileges of western, 'universal' knowledges and institutions' (Radcliffe and Radhuber 2020: 2).The 'proof of the [decolonial] pudding' is arguably identified in how research is analysed and presented (Hitchings and Latham 2020: 392).As such, the sharing of the deconstruction, evaluation and critical reflection on methodology is critical to foster knowledge exchange between scholars, practitioners, and respondents for more equitable research processes and outcomes.In outlining unplanned methodological intervention, the research aims to contribute to the growing literatures on both researching remotely during the pandemic concurrently with decolonising methodologies in energy studies.
To build a collaborative knowledge, Smith (2013) argues that decolonising methodology emphasises on the context in which research is designed and its subsequent implication on participants and broader community networks.Initial consultations with project NGO partners and facility manager, who acted as research collaborators, were to be centred around the logistics of conducting research remotely.The authors were careful on building a relationship to avoid extractive reputation.It was collectively decided to pursue the project in a workshop format, that prioritises skill exchange between participants and researchers, as well as providing official letters of achievement at the end of the process as an acknowledgement of the skill development and participation in general.This decentring of the researcher resulted in the adaption of research objectives and process to respond to the priorities of the participants (Radcliffe 2020).
Decolonial methods can be seen as 'communication strategies' that enable participants to examine their personal, societal, and institutional networks through means such as storytelling, challenging dominant forms of knowledge building (Chinn 2007: 6-7).The remote working presented a challenge in engaging meaningfully with participants and reading cues such as body language, to gain informed consent.To address this, animations that represented participants were created to 'tell the story' of the research, explaining consent, process, reasoning, and research promises in the local language (see Sec. 3).Furthermore, the animations were framed as a 'request for participation' , giving participants agency in deciding if the workshop and its outputs would be of value to them.These videos and posters enabled us to address the 'mismatch' between the timelines and expectations of participants and researchers (Banks et al. 2013), with social workers on site taking time to explain to various community members, and for participants to view the information when convenient, and as many times needed, before signing consent forms.This reflexive approach is an opportunity to reflect on the 'fetishes of consent' such as the written form of consent agreements on university headed paper, that present the charade of neutrality (Wynn and Israel 2018) but can reinforce unequal power relations between participant and researcher.Such methods do slow down the research design process prior to data collection, and but establishes meaningful and informed participation in the research.Hence the method is applicable beyond the studies that were forced to move online due to COVID-19.The adjustments made resulted in a critical re-evaluation by the authors on how we conduct research in the future and is of relevance to all forms of research-particularly those that negotiate unequal power dynamics between participants in the research process.
In this study, a number of stakeholders were involved with varying levels of power in defining the research process.The NGO partners were important in brokering the relationship between participants and the researchers and are also the connection between housing beneficiaries and government actors.They have established high levels of trust and influence between all involved stakeholders, and we were heavily reliant on their expertise, intuition and knowledge in the research design.The NGO partners were equal research partners from the outset and they were funded by the Research Council.In Pickwick, the facility management (Miradi) social workers were a key contact in understanding the capacities of participants to participate, and we were reliant on them for the more intimate interactions with participants (such as checking in on their welfare), which we ourselves could not facilitate due to the remote nature of the study.It was through their knowledge, that we came to realise that the women in Pickwick in Cape Town had limited access to camera phones, unlike the SRA participants in Mumbai.It was suggested that a young resident, who has grown up with the women since their time in the informal settlement could be the assistant filmmaker in Pickwick and be paid for this role which helped to broaden their portfolio for future work.
For both studies, the authors' longstanding relationships with beneficiaries and the NGOs provided a foundation of trust and familiarity which contributed to participants incentives to participate.It was through a slow process of negotiation and listening that we could understand what in the research process would bring value to participants, which can be summarised as: (1) clear communication of the exact number of hours of participation, (2) payment for labour, (3) official letter of recognition to aid future job prospects (acknowledgement of participation), (4) space to be listened to and respected in group and individual settings.Aside from these practical investments in participants, we hope the study fosters inter-stakeholder relationships where policy makers are able to see inside their policies through the films, and participants are able to articulate how their spaces may or may not contribute to the processes and experiences of their everyday life.In order to facilitate this, the compilation film was screened in policy workshop in Johannesburg in May 2022.Ultimately, the 'politics of inviting' lie between what we the researchers, the NGO's and the participants bring to the table, and how these are co-articulated and relationally re-ordered (Lindstrom and Stahl 2016: 194), resulting in a mutually constructed process that has value to both researchers and participants.
The process of decolonising research transforms researchers for participants (Datta 2018: 1-2), and the need for increased trust in partners and participants in a remote study, established deeper relationships, cultivating empathy within the researchers to respond meaningfully to the concerns and realities of participants.The myth of neutrality resulting in more reliable outcomes (Given 2012;Saltelli et al. 2020) was firmly unsettled through this collaborative methodological process.Table 1 describes our research process and how it was transformed due to the COVID-19 pandemic.
Moving the study online due to the pandemic facilitated deeper reflections with regards to decolonising research practice and provided stimulus to consider shifting positionalities, and the shifting of the space of the research.The research space exists with the researcher however in this case, the space shifted or extended with the participants.Despite these gains in new reflections concerning the co-production of research and the decolonisation of practice, feminist ethical considerations were compromised.Maria Lugones' (1987) call for 'loving perception' for example, requires the abandoning of distance and to engage with participants with our whole selves, rendering the production of knowledge political and personal, as connections are made between research, activism, and women's everyday lives.These considerations of ethics of care are reliant on intimacy to provide support to participants and research assistants, some of which was inevitably lost when the study went online and digital barriers were created.

Case studies
Two case studies were selected for the study: Slum Rehabilitation Authority (SRA) housing in Mumbai and Pickwick transitional housing in Cape Town.Pickwick transitional housing in Cape Town, South Africa is a change from South Africa's typically output focused housing delivery models, with its social programme and management systems that are intended to help people 'transition' into formal living by addressing the social and material barriers to housing stability.Pickwick houses nineteen households who have survived multiple displacements from their prior informal settlement home.The accommodation is a two-storey building with a guarded entrance gate, a large communal yard, outdoor washing area, two shared kitchens and shared bathrooms (see Figure 1).It is comprised of 39 units that range from 10-15 m 2 depending on household size, and residents pay a low monthly rent which includes access to uncapped gas, water and electricity.In principle Pickwick is considered 'temporary' , in that the housing is explicitly designed as a short-term intervention to address the structural causes of housing insecurity, with beneficiaries expected to eventually 'transition' into more permanent housing, thereby making way for other short-term tenants but in reality, the residents live there long term.
Natwar Parekh Colony in Mumbai, India, is commissioned by the Mumbai Metropolitan Region Development Authority (MRDA) under the Slum Rehabilitation Act 1995 to rehabilitate eligible slum dwellers and resettle (infrastructural) Project Affected Persons (PAP).The slum rehabilitation policy is housing policy where the private sector is incentivised to participate in the redevelopment of slum communities.This model allows for slum dwellers to voluntarily opt for redevelopment by forming co-operatives through a developer for construction of the redevelopment units.The slum dweller gets a formal tenure-ship to the allocated rehabilitation unit, with no upfront costs, giving a sense of stability.Economically, the participatory basis of SRA doubly benefits the Mumbai government: by opening the landlocked areas for development with minimal intervention and provides opportunity to rapidly close the housing deficit.Natwar Parekh Colony represents linear SRA housing typology with gallery access to tenements (see Figure 2).High-rise buildings are lined in dense rows and there is no communal outdoor space.The housing units are accessed through a single loaded corridor and all unit open to a public access corridor.A unit is usually of 24 m 2 and often home to 6 or 7 people.The residents need to pay for maintenance fees and energy bills.
The research participant recruitment and research design were done in collaboration with the NGO partners in Mumbai and Cape Town (see Sec. 2.2).Apart from the assistant filmmaker in Pickwick, the research participants were all women: young adult women with families, who had lived in the housing for at least 2 years.There were 5 research participants in Pickwick (Interviews A-E) and 8 participants in SRA (Interviews 1-8).Due to the confidentiality of the data, we do not present a table on the participant demographics, but they all represent low-income households, have all lived in housing for at least two years and had no intention of moving out at the time of the study.Some participants had academic degrees but only two women were in employment in each housing.
Two workshops, one before the filming exercise and the second after the films were submitted by the participants, were conducted with the NGO partners and the research participants in both locations.Consent of participation were taken after briefing the participants about the research objectives and the expected outcomes.The theme of the workshops was: '5 women-5 days-5 moments in Pickwick/Natwar Parekh' .The research participants were requested to film 30 second video clips of their daily practices, recorded through a period of five days, at a three-hour time interval starting from 6:00 clock in the morning (i.e. from 6:00 to 20:00 hours).The first workshop focussed in training the participants on the method of filming the video clips to ensure that no face or other personal identification features were visible in the films.The participants' films focused on their daily activities like cooking, cleaning, childrearing and entertainment practices, and the use of TV, kitchen appliances, and the energy infrastructure (LPG, electricity).The participants' films were then collected through an online data storage platform, where the participant had the right to upload the films that they deemed fit.The NGO partners, and the assistant filmmaker in Pickwick, co-ordinated uploading the films.
In the second workshop, the researchers asked semi-structured questions from the participants, based on the films observed.The questions were same for all the participants, such as: How much time spent on cooking per day (time per session)?How do you ventilate you room?What do you watch on TV: soaps, news, educative programs?
Due to COVID-19 pandemic all workshops took place through online video conferencing platform (zoom), hosted by the NGO partners who were in charge of bringing the participants together.All the workshops and research interviews were recorded with prior consent.The length of each workshop was 60 minutes.The research process was explained visually to all participants through a printed poster (distributed prior to the first workshop) (see Figure 3) and in a sample film.

Participatory filmmaking in this context
The participatory action research approach adopted in this study differs from conventional research through its' reflective cycle, where participants collect and analyse data (Baum, MacDougall, and Smith 2006).It pays specific attention to power relationships between the researcher and the researched and gives agency, in this case through self-presentation in film, to the researched who are not considered objects or stakeholders but become researcher partners.Visual ethnography and participatory video have been advocated as research methods in social sciences (Lenette et al. 2020;Pink 2021) and development projects (Asadullah and Muniz 2015) but the definitions of what is meant by participation and research partners differ.
In this research, we consider research partners to be the NGO partners, social workers and the assistant filmmaker in Pickwick.The NGOs were research partners from the outset of the project, they determined the research objectives, data collection and participant recruitment.It would have not been possible to conduct the research without their involvement.Lenette, Cox, and Brough (2015) argue that participatory video has the political potential to contribute to creating counter-narratives that challenge leadership while recognising that it would be naïve to assume decision-makers would come across the filing and know what to do with it.Lenette, Cox, and Brough (2015) argue for involving decision-makers from the outset.In this study, the City of Cape Town was a research partner from the beginning of the project.
Participatory filmmaking in this context has similarities to an autoethnographic research approach where participants turn the camera to themselves.The research participants filmed themselves to share their perspectives and by doing so, they offered us as the researchers an opportunity to identify wider socio-cultural elements that shape those experiences.The instructions for filming, including timings and length of the films, were given by the researchers but after that the researchers did not guide the filming in any way.The participants decided themselves what they wanted to film and some of them decided to add a voiceover, such as commenting on their 'favourite place' in Pickwick (e.g. the communal bathroom with showers, the wifi-spot with a view over mountains or their private room) and the 'saddest place' (e.g. the space by the bins where children like to play).The participants commented on the films in the screenings and contributed to the analysis as 'co-researchers' .Film data can be further counted and coded in order to be transformed into research data that can then be processed or analysed (Goldman andMcDermott 2009, Snell 2011).The films were viewed and analysed by the researches, including quantitative observations such as length and location of the shots.The film data material was also reflected upon by the participants (NGOs, social workers and the participants who did the filming) at multiple points in the process, which were less intrusive opportunities to ask interview questions.The process culminated in a screening of the compiled films and the distribution of the letters of acknowledgements of the participation.
Jewitt (2012) identifies three features of film data that underpin its potential for social science research: (1) the ability to capture temporal structures and real-time sequential records, (2) creating fine-grained multimodal record, and (3) its durability, malleability and share-ability.The third point on malleability was adapted as the final part of the analysis and the participants' films were thematically compiled into two short films: 'A Moment in Pickwick' (3:49 mins) and 'A Moment in Natwar Parekh' (3:26 mins).This was done by resampling key scenes from the films, following a practice at certain time of the day (e.g.cooking breakfast, making bed in the morning), in order to compare the practices and to generate a sequence of events of a typical day.This is what Goldman (2009) calls 'chronological verisimilitude' , meaning the use of film data to represent the order of events, not necessarily as a chronological order but to enable the viewer to comprehend events 'in sync with the meaning of events' .Split screen was used to enable simultaneous representation and comparison of practices in different households.There was a deliberate attempt to avoid 'aesthetical' editing, making the films to 'look good' .However, aesthetics are likely to have guided the participants' film narratives and although it was not asked in the interviews, it would be useful to know whether they took several cuts before submitting a film in order to make it more representative-this 'looking good' factor is not usually considered in participatory filmmaking but it is likely to influence it although people's perception of 'looks good' is likely to differ.
When the editing was done by the researchers, a concern of participatory video making the risk to misrepresent the participants' views and imbalances of power of stakeholders (Lenette et al. 2020).In this research however, the main unit of analysis were the participants' films that were not subject to any post-production and the final stage compilation films were made for policy workshops, and it was seen that co-editing as a group would have led to negotiated outcomes and long process with different members of the group entering the process with different ideas and expectations at different times, posing a risk that a compilation film would have represented what could be agreed by a diverse group of people, rather than an output that reflected accurately the collective position, as described by Yap (2021).
However, in order to overcome any risks of misrepresentation, 'rough cuts' were screened for the research participants.Discussions with the NGO partners, previous engagement with the research participants and pre-pandemic fieldwork helped to ensure that the analysis was not misrepresentative or open to assumptions.
Finally, we want to acknowledge our positionality as researchers.We acknowledge the power asymmetry that exists between the various actors including ourselves within the research process.We are conscious of our positioning in the process, and the participatory filmmaking did offer us new and unexpected perspectives, challenging our assumptions and ways of knowing.Technology itself changed power relations in the research process: some research participants were clearly more in control of the technology and were more knowledgeable about video making than us.Technical competence or lack of methodological understanding was not a barrier.Time commitment to do the filming was not an issue and was not commented on by any of the participants.
It should also be noted that knowing the researchers was crucial in the NGOs and the residents' decision to participate to the project.In addition to this, it became clear in Pickwick that the process should involve also the facilities management company and social workers on site, whose participation was crucial for recruitment and organising the filming workshops.One of the researchers had conducted pre-pandemic fieldwork in Pickwick and her knowing some to the residents helped to get them to commit to the project.Our experience supports the observation by Lenette et al. (2020) that participatory video research has implications for ethics committees that traditionally tend to prefer distance to participants.

Clustering of practices
The total number of films recorded by the Pickwick residents were 75, with an average duration of 22.45 seconds.119 films were recorded by the Mumbai SRA residents, with an average duration of 35.25 seconds.The films were coded according to (i) practices and length observed in the film; (ii) reliance on energy infrastructure; (iii) location of practices.The daily practices were coded into four types (i) cooking, (ii) maintenance activities, (iii) recreation and religious activities, and (iv) childcare.The cooking practice involved the preparation of cooking near the stove, and the actual act of cooking.The maintenance activities included practices of cleaning like washing clothes (by hand and in a washing machine), doing the dishes, sweeping and mopping of floor, sorting and cutting of raw vegetables to be stored for future cooking, collection of water and watering of plants.Sowing, watching television, handcrafts, praying, painting or artwork and eating with social network were described as recreational activities.Childcare comprised all activities related to childrearing including bathing, feeding and helping with schoolwork.Figure 4 shows the key practices films by the participants in SRA and in Pickwick.Figure 5 shows the locations where the filming took place and how most domestic practices take place indoors: 95% of the participants' films were made inside tenements in SRA and 59% in the private rooms in Pickwick.
Screen shots from the participatory filmmaking are shown in Figure 6 (SRA) and Figure 7 (Pickwick), showing key practices and when they were filmed.The films were very effective in capturing the material context in homes: heterogeneity of the conditions, daylight, the presence and use of domestic technology, and how for example cooking overlaps with other practices such as watching television or childrearing.It would have been very difficult to capture the information in interviews: the films were literally worth thousand words.

Cooking
The films show that cooking is the most persistent practice when moving from informal to formal housing and it is the most time-consuming activity.In SRA housing, women still cook 3-4 times per day with a duration of 1.5-2 hours per cooking activity, using LPG as the cooking fuel.Figure 8 shows screen shots from the participants films: daily preparations for cooking in the morning (8a) and cooking (8b) in the afternoon.
The participants in SRA cook traditional food, always from fresh, and there is little variation in cooking practices regardless of the participant's religion, age or household type.The films commonly showed a cookstove but also special appliances like mixer grinder.No help from men, children or extended families was visible in any of the films.Some participants put in approximately 56 hours of work just on cooking per week: Fridge is often seen as 'time machine' in Western literature (Shove and Southerton 2000).Most SRA participants had a fridge but they use it to store raw cooking materials rather than cooked food.They would only refrigerate raw vegetables, milk, water, meat or fish, or pre-cut the vegetables and refrigerate them to save time during cooking sessions.In the films, all meals were always prepared fresh: I don't store cooked food as nobody wants to eat food which is not freshly cooked.It is also waste of cooking gas re-heating stale food.It is an unnecessary wastage.(Interview 4) My husband does not like stale food.I and my daughter can eat refrigerated food.But my in-laws will not eat that food.(Interview 8) I have fridge but no one will eat stale food.They want warm freshly cooked food.(Interview 5) Previous studies in rural India (e.g.Wilhite 2008;2018) have documented prejudices against refrigerated food and the storing and reheating of cooked meals as inferior to fresh meals.Less than 30% of Indian families have a fridge (The Indian Times 2017) and the ownership is small even in middleand higher income groups.This study suggests similar attitudes in urban Mumbai-despite steady power supply and women having little assistance in household tasks.
Cooking was the practice that differed the most between the two cases.In Pickwick, the women make eggs, pasta, sandwiches and tea, and use canned products that are heated up in electric pots or 2-plate electric stoves in their rooms.There is little use of meat, vegetables or other fresh ingredients.There was also less frequency in cooking: 'I do not like cooking during the day, if I cook it is for the evening' (Interview C) and aspiration to cook less often: 'I like to cook twice, I do not have a fridge, that is why I cook twice' (Interview A).Unlike in Mumbai, no reservations were expressed towards refrigerated food: 'If I had a lot of money, I would buy me a fridge' (Interview A). Figure 9 shows cooking practices in Pickwick.Electric stoves and electric kettles (heatable pot that includes heating and does not need a stove, see Figure 9a) were used for cooking inside the rooms (see Figure 9b).
An important work in feminist analysis of domestic technology is Cowan's (1989) book More Work for Mother which analyses the time saving benefits of household appliances in post-War US.Cowan (1989) shows that time savings anticipated from household technology and energy access did not fully materialise in post-war America.The time saved in cooking or cleaning was simply filled with other household tasks.Cowan (1989) argues that while energy infrastructure and new household technology were marketed with the narrative that they would free women from household chores and increase their comfort at home, in some cases it increased the housekeeping standards, such as cleanliness or more elaborate cooking expectations.At the same time, household help became more expensive: women who used to have household help or support from extended family found themselves in a situation where they had to do all domestic chores themselves.Wilhite (2018) argues that Cowan's questions on More Work for Mother are still highly relevant in the Global South.Our findings suggest that the time spent on cooking and household chores, especially in Mumbai, echo with the work of Cowan (1989): energy access and the use of domestic technology does not necessarily reduce the overall time spent on household chores.The films show that the women use the time saved on taking up more household chores or increase the standards of good housekeeping and cooking: they still stay at home.However, Mallett (2004) points out that gender studies leave little room for 'identifying the positive attributes related to homemaking, like place attachment, personal satisfaction, and meaning' (Mallett 2004: 77).It was noticeable how the women in SRA took pride in their cooking: looking at it from utilitarian and time use perspective would be missing the point.

'Everything I do is in my room'-privacy as privilege
Despite the spacious communal kitchen in Pickwick, the films show how the participants prefer to improvise private 'kitchenettes' in their own rooms, even if they have little furniture and no water tap in the room.They prefer to cook and eat in their own room, even if it means cooking on the floor and without furniture: 'We eat in our own space.We used to cook in the kitchen when we were seven, now gas runs out, now people start to cook in their own kitchen stoves, we cook in our rooms… I would like kitchen in my room, you are there to supervise your pot' .(Interview C) The communal kitchen in Pickwick is used for washing dishes and occasionally to cook on LPG stove but it is not considered as a place to linger: 'I like to cook in my room, I cannot go up and down every time, then I have to stay in the kitchen the whole time' (Interview A).Cooking in the communal kitchen is considered inconvenient: In communal kitchen you need to carry everything with you (Interview C).Also in SRA housing there was little interest in having a communal kitchen 'everyone's food is different' .
In both cases, the myth of communal space as functional and desirable for the poor in slum upgrading process was firmly disproven, as the films expressed privacy as a sign of prosperity, as a privilege usually reserved for the rich.Own room is where the women like to spend most time in Pickwick: Everything I do is in my room … if I do not go to hospital or shops I mainly stay in my room, watching videos (Interview B), and it is where the main future aspirations are related to: I would prefer to refurbish my room.I really want to give my room a makeover, for me and my daughter (Interview C).

Ventilation and security
The rooms in Pickwick and SRA are very small and have a high occupancy rate-in SRA up to 7 people live in a room of 25 m 2 .Without cross ventilation and adequate daylight levels the households need to resort to electric lights and fans just to keep the space habitable.In Mumbai case study, doors and windows of single-aspect tenements only open onto to the public

Domestic technology and identity
Figure 10 shows the domestic technology filmed by the participants.The size of the text is proportional to the frequency seen in the films (Figure 10a) and television is the most used appliance, especially in Pickwick (Figure 10b).
Women's media watching habits have gone through radical changes in India (Namrata and Kakade 2014;Pugalendhi 2015;Bhatt and Singh 2017).Even if the participants in SRA said they watch very little television or do not own one, Debnath et al. (2019) have shown how in the transition from informal settlements to SRA, time spent on watching television has doubled.In Pickwick television is the background for most practices (cooking, eating, cleaning, childrearing) and it is watched throughout the day: In these two contexts domestic technology is sign of wealth and settled status and the films show that it is important in homemaking.Gram-Hanssen and Darby (2018) have described how femininity and masculinity are inevitably encoded in technological artefacts at home and in domestic practices and skills.They argue that technology is traditionally seen as a masculine-coded artefact, whereas the home, homemaking and everyday practices, are dominantly considered female.In Western literature, 'the home' is seen as a medium of self-representation, a sensory environment that contributes towards the recreation and extension of gendered identity (Pink 2004).In these two cases, however, the dichotomy of seeing 'the home' as feminine and 'technology' as masculine is more blurred.The institutional and monotonous character of transitional housing can be read as a 'masculine' technical solution to housing crisis, whereas the participants' films showed us how domestic technology, such as a table fan or a television, are important in 'feminine' home-making and used to personalise and distinguish private spaces. of empowerment: (1) government and policy makers (building the infrastructure through which beneficiaries can exercise their empowerment), (2) NGOs (building the capacity of the urban poor through facilitating their capacity for agency), and (3) beneficiaries (understanding and envisioning their own empowerment, and having the agency to take strategic action leading to specific outcomes).By using filmmaking, our research intends to build the 'awareness of one's own potentials and the confidence to develop them' (Friedmann 1992).Lenette et al. (2020) argue that when protagonists are placed as filmmakers and share their first-hand experience, they can 'view themselves as agents of social change who take initiatives, make decisions and share opinions as counter-narratives to dominant portrayals of passive, vulnerable and helpless victims' .In this study, the participants expressed their aspirations (Figure 11) and described how the filmmaking made them see their time use (Figure 12), what type of skills they already have, and how it made their work 'visible': I can make videos of my recipes now.My family members inspire me to take up entrepreneurship in food catering, but it is difficult to take time out of my daily schedule for such activities.(Interview 1) This video making helped me with the spend my time.(Interview 8)  The films aimed to not just understand the ambitions of empowerment from the perspective of the women, but to also create a space for reflection and the imagining of new futures.Empowerment in this sense, echoes Harvey's (2003) sentiment in A Right to the City as being far more than the right to access what exists, but a right to change it.

Conclusions
The COVID-19 pandemic forced a reassessment of research objectives and methods, as we navigated the unforeseen situation of conducting research remotely, inadvertently applying decolonial practices to centre participants in the construction of the research.
The paper adopted dual approach.First, to study gendered energy transitions in low income housing in urban India and South Africa.Second, the study argues for participatory filmmaking as a method in energy studies.The films revealed that access to formal housing and electricity grid have led to improvements in the women's welfare and wellbeing, showing the development of personal leisure activities and the importance of privacy.Energy infrastructure provides the women with comfort, whether through warm meals, television as childcare device, or access to hot showers: ' … for 25 years I have not seen hot water'.The importance of 'family-scapes' (Appadurai 1994) was observed in both contexts: the role of grandmothers as the main carer for children (Pickwick) and resilience during the COVID-19 pandemic and the first lockdown in India when some women returned to their own family home to be supported by their extended family.However, there is also pressure on women to provide for the whole 'family-scape' and the in-laws were also the first people the women wanted to show their films to show them how busy they are the whole day.In Mumbai, the women still cook 3-4 times per day, every cooking moment taking 1-2 hours, amounting to women putting in 56 hours of (unpaid) work per week.
This study has demonstrated through two case studies that the simplistic assumption that energy access/usage has no relationship to gender mainstreaming or empowerment is problematic.The relationship is neither spurious nor linear and is embedded in the definition of a good wife/mother.In India a good wife/mother is a subaltern who has uncomplaining acceptance to the life defined by her husband or in-laws, has higher tolerance to extreme situations and puts the children and family before her needs, often at the cost of her comfort or health (Spivak 1988;Cherian 2008).The women in the surveyed Mumbai slum rehabilitation housing, though having access to 24 hours electricity, were restricted by socio-cultural notions of a 'good wife and mother' , limiting their agency to enjoy the benefits of electricity.While they were aware of the benefits of a refrigerator, they still preferred freshly cooked meals and considered pre-cooked refrigerated food as 'stale' .The cultural obligation of women towards the social reproduction inside the home and the compelling need to adhere to be 'a good wife/mother' rendered them as shock absorbers.Similar evidences were reported by Malik et al. (2020) in their study on the gender dimensions of thermal comfort in Mumbai's slum rehabilitation houses.They found that women had higher adaptive saturation for comfort temperatures then the men in the family.The social conditioning that women who mostly stay indoors should be responsible for managing the electricity bills, limited their usage of electricity.Hence, contrary to the western belief that electricity access empowers women, here the empowerment was primarily defined by the socio-cultural norms.
In line with decolonial practices, the research participation was extended beyond method and into the research design (Radcliffe 2020), as the participants were invited to reflect on the data through interviews, participating in the analysis.Filmmaking gave the participants autonomy and control over showing what they consider important in relation to the research objectives, ensuring dignity, and mitigating the possibility of the removed western researcher fetishizing the aesthetic of poverty.The films captured the heterogeneity of the material arrangements and allowed the participants use their whole body to show the ways they experience everyday life at home.For us as researchers, the films gave 'non-miserablist' perspective to daily life in these low-income communities.For the women, the films made their work visible.We argue that empowerment, also in the field of energy studies, is and should be a choice-'it refers to the expansion in people's ability to make strategic life choices in a context where this ability was previously denied to them' (Kabeer 1999).This visibility can contribute to women's self-awareness, capacity and ultimately contribute to people's agency in exercising choice to create new futures that are meaningful to them and contribute to their own visions of empowerment.

Figure 3 .
Figure 3. Printed poster visually explaining the research process to participants in Pickwick.

Figure 4 .
Figure 4. Key practices filmed by the participants in Sra (Mumbai) and Pickwick (cape town).

Figure 6 .
Figure 6.Screen shots from the participatory filmmaking in Pickwick (cape town), organised by the type of practice and time of filming.

Figure 7 .
Figure 7. Screen shots from the participatory filmmaking in Sra (Mumbai), organised by the type of practice and time of filming.

Figure 8 .
Figure 8. Screen shots from the participatory filmmaking in Sra (Mumbai), showing: (a) daily preparations for cooking (top), and (b) cooking in the afternoon (bottom), the quotations from the research interviews.

Figure 9 .
Figure 9. (a) typical electric kettle used by the women in Pickwick (cape town) (top), and (b) screen shots from the participatory filmmaking, showing the participants' daily cooking practices in their private rooms (bottom).
corridor, and due to cultural conventions, risks of overlooking and lack of privacy, windows and doors are kept closed even when cooking: I never open the windows due to lack of privacy.(Interview 5) I only open windows when I feel extremely uncomfortable due to heat, otherwise it is closed.I do not keep the window open at night.(Interview 1) At night to avoid mosquitoes and rats I keep in the window closed.I also close the window while cooking.(Interview 4) By contrast, Pickwick has 24-hour security and fencing on site, which enabled women to feel comfortable and secure in keeping windows and doors open during both the day and night.This ensures natural ventilation and healthier indoor conditions.It is the security infrastructure in this context (fence, controlled access) that enables better thermal comfort for women.I keep windows open during the day and during the night.Room is very hot in summer … I leave the windows open but at night I close the curtains.(Interview E) In the morning I start my television at 6 for the news and then keep it on until my soaps start at 9.30.(Interview C) I watch TV until 1 o'clock in night-time, I start around 6am, I watch the news, my soapies … from 3 o'clock I watch TV.(Interview A) My grandson is paralysed, he is in wheelchair, it is TV day and night for him.(Interview D)

Figure 10 .
Figure 10.(a) domestic technology observed in the participants' films in cape town and Mumbai, and (b) screen shot from one participant's film in Pickwick, the quotation from the research interview.

Figure 11 .
Figure 11.aspirations expressed by the research participants in Pickwick (cape town), screen shots from the participants' films and the quotations from the research interviews.

Figure 12 .
Figure 12. the Sra participants (Mumbai) commenting on the how the films made their work visible, screen shots from the participants' films and the quotations from the research interviews.

Table 1 .
collaborative methodological process adopted in the research.