Theorizing the Continuities between Marriage and Sex Work in the Experience of Female Sex Workers in Pune, Maharashtra

: Marriage is near-universal in India, where most cisgender women sex workers have been married at some point in their lives, while also navigating responsibilities to family and children. In this paper, we explore how cisgender women sex workers in Pune, in the Indian state of Maharashtra, experience continuities between sex work and marriage, while navigating an ideological landscape where sex work and marriage are positioned as opposites. Returning to feminist theoretical models that highlight the economic underpinnings of marriage, we outline three arenas in the Indian context where marriage and sex work overlap rather than remaining opposed and separate entities: (a) migration, (b) attributions of respect and stigma, coded through symbols of marriage and sexual availability, and (c) building and dissolving kinship networks that contest the primacy of biological or affinal kin. In each of these realms the distinction between marriage and sex work is a fraught and contested issue, and the roles of wife, mother, and sex worker can shade into one another based on context. We then examine how three women navigate these contradictions, arguing that focusing on kinship and marriage can circumvent the limitations of the choice versus coercion paradigm that structures current debates on sex work. empirical


Introduction
I was married soon after I reached puberty, but my husband was very abusive, so I left him. I moved to Delhi with the help of a friend and joined sex work. I never saw my husband again, though I still wear my mangalsutra 1 and toe ring. It helps me with my work. It keeps me safe from unwanted men and people respect me. I sent money back home regularly to get my siblings educated and married. My children are also married now.
---K, age 38, Budhwar Peth, Pune K shared this reflection with Sutapa over a cup of tea in her room in Pune's red-light district, as the monsoon rain pummeled the streets outside. 2 Since leaving her husband, she had lived all over the country doing sex work: in Sonagachi, Mumbai, Delhi, Hyderabad, Mysore, Kakinada, and Bangalore. She had even returned home briefly to do agricultural work and domestic work, before returning to sex work in Pune. When Sutapa met her, she worked as a peer educator at Saheli Sangh (more commonly known as Saheli), a community-based organisation advocating for sex worker rights. 3 While K's "real" marriage did not provide much support or stability, the symbol of marriage, namely the mangalsutra, gave her a sense of respect and safety while doing sex work and other forms of work for over two decades.
K's use of the mangalsutra is extremely common amongst cisgender women who do sex work in India. 4 Anthropologist Svati Shah has documented in Mumbai how women involved in sexual commerce wore mangalsutras regardless of religion (Shah 2014, p. 107). Other researchers have shown that symbols of marriage and domesticity help sex workers avoid the stigma of sex work in a middle-class neighborhood (Guha 2018a;Dasgupta 2019), and that appearing to be Hindu, married with children, and of higher caste can partially "override the stigma associated with sex work" (Hui 2017). The prominent sex worker rights activist Meena Seshu, commented at the sex worker rights conference in 2003: "Why is it that all women in prostitution wear the mangalsutra? Is it because they are all married? Or is it because they are showing that they have a license to have sex?" (cited in Shah 2014, p. 220-221n28).
What do we make of the ubiquitous use of symbols of marriage amongst women doing sex work, or pursuing work more generally? What is its significance for women who, paradoxically, have been driven to work by the dissolution of their marriages? Some feminists have interpreted these practices of wearing symbols of marriage as an unfulfilled aspiration for the domesticity of heterosexual marriage. 5 The problem with this reading is that it implicitly treats "wives" as the original point of reference, with sex workers as aspirants or pretenders. In India, however, where marriage is a near-universal experience 6 amongst cisgender women, most women in sex work are not imitating wives, they are wives. Sex worker collectives in India repeatedly point out the continuities between the labor of a wife, mother, and sex worker. While many have experienced a dissolution of ties with their natal family or husband's family, they remain tied into the marriage economy, using their earnings to pay dowry for the marriages of family members, sending remittances to their husband's family, or paying for the care of their children. 7 Others are "flying sex workers" who might be married while doing parttime sex work (Kotiswaran 2006, p. 2). The respondents to this study had critical, reflexive relationships with their identity as "wives", and were skeptical of the short-lived "safety" provided by marriage. While she wore the mangalsutra, K did not consider her marriage to be her primary locus of identity. She highlighted her identity as a worker, as a sibling, a mother, and a community member with Saheli. Exchanges of sexual labor, she pointed out, are not limited to settings of sex work or marriage. Women in sex work often share that they have experienced sexual demands from other employers, whether working in construction University and was awarded her doctoral degree in May 2017. The case studies of female sex worker were collected after securing verbal consent of women in sex work and with support of the Saheli Sangh, a sex worker collective, , which works in the red-light district of Pune city. Saheli Sangh has been thoroughly acknowledged in her doctoral work and beyond. 3 Saheli Sangh works with two lodges, 100 non-brothel-based sex workers (NBBSWs), and approximately 2000 women in sex work living in the Red Light Area of Pune to enhance and enable greater levels of self-protection among sex workers through a sense of togetherness, collective action, and most importantly creating an identity. Saheli is headed by Tejaswi Sevekari, Executive Director. 4 This paper limits itself to an analysis of cisgender women's experience of sex work, and not those of transgender women, male sex workers, and other sexual and gender minorities. In the remainder of the paper, we use the term "woman" to refer to a cisgender woman. However, a critical question for further research is how marriage economies, migration, and informal labor structure sex work amongst other gendered communities. 5 For example, see Ghosh 2017, p. 197. 6 The National Family Health Survey-IV shows that by the age of 45-49, only 1% of women and 2% of men have never been married, and less than .5% of them report being divorced (IIPS 2017, p. 155). 7 Nalini Jameela, the noted author, activist, and sex worker from Kerala, has written extensively on the parallels between marriage and sex work. She herself started sex work after her husband's death, when her mother-in-law started demanding money in exchange for taking care of Jameela's children (cited in Kotiswaran 2008, p. 225).
sites, brick kilns, domestic work, and farming (Kotiswaran 2008, p. xiii). K's use of the mangalsutra offers a powerful illustration of how, as women move across different kinds of labor, vocabularies of respect and stigma, coded as marriage or sexual availability, respectively, continue to structure their experience of work.
In this paper, we examine how women experienced continuities between marriage and sex work while inhabiting an ideological and material landscape where sex work and marriage were positioned as opposites, and the opposition was maintained by regulatory regimes that heavily criminalized sex work and endorsed marriage. We argue that focusing on kinship and marriage economies can circumvent the limitations of the consentversus-coercion paradigm that structures many current debates on sex work. To theorize the continuities between sex work and marriage, we return to three feminist theoretical models highlighting the economic underpinnings of marriage as a system that formalizes gendered divisions of labor, particularly sexual labor. In the second section, we apply this theoretical framework to the empirical realities of how sex work takes place in Budhwar Peth, Pune's red-light area. We identify three arenas where the continuities between marriage and sex work become especially visible: (a) in the process of migration; (b) when navigating respect and stigma, often coded as sexual chastity or availability, respectively; and (c) by building and sustaining kinship networks that contested the primacy of biological and affinal kin. In each of these realms, the distinction between marriage and sex work is a fraught and contested issue, and the roles of wife, mother, and sex worker can shade into one another based on context. We illustrate these continuities by sharing the life narratives of three women. The life narratives analyzed here were collected as part of the doctoral research work of Sutapa Majumdar, the co-author of this paper. We conclude by sharing the implications of our analysis.

The Economic Underpinnings of Marriage
The relation between marriage and sex work has been theorized extensively by feminists of almost every tradition. We build a dialogue between three formal models in particular, which are especially suited to the Indian context: materialist feminism as developed by Italian feminists in the "Wages for Housework" movement of the 1970s, queer theorist Gayle Rubin's theory of marriage as a "traffic in women" (Rubin 1975), and historian Uma Chakravarti's elaboration of the concept of Brahmanical Patriarchy (Chakravarti 1993). Each of these models denaturalizes and complicates the category of "wife", pointing to how heterosexual marriage and kinship reproduces divisions of labor that map onto gendered and caste-marked identities. We then demonstrate how this these formal models are helpful for addressing the empirical realities of sex work in the Indian context, before turning to the life narratives of specific women.
(1) Materialist feminism Socialist and Marxist feminists have long debated the place of housework and sexual labor in the Marxist theory of capitalist production. These theorists sought to recognize the labor done within marriage, demolishing the ideological opposition between the private realm of the family and the public setting of the factory so central to orthodox Marxist theory. 8 Feminists from the "Wages for Housework" movement of the 1970s, such as Mariarosa Dalla Costa, Leopoldina Fortunati ([1981] 1995), and Sylvia Federici (1974, challenged the productivist bias in Marxist theory, which recognized the labor of the male wage worker but ignored unwaged labor such as that performed by housewives (summarized in Meehan and Strauss 2015, p. 7-9; see also Mitchell, Marston, and Katz 2003). In The Arcane of Reproduction (Fortunati [1981] 1995), Leopoldina Fortunati proposed a formal model that illustrated how the unpaid labor of housewives constituted the invisible "arcane" part of (male) production, since it was an essential part of reproducing the labor power of the male factory worker. In this perspective, gender roles subsidized the capitalist production process, as heterosexual marriage gave husbands a monopoly over the unwaged labor of housewives. The sexual labor of prostitution 9 complemented the labor of housewives by making up for a deficit in marital sexual relations.
The Wages for Housework movement is a key theoretical influence for social reproduction theory, which expands the limiting category of "housewife" (as theorized in Fortunati's model) to analyze the value of unwaged and undervalued labor more generally (Meehan and Strauss 2015, pp. 8-9). From the 1990s onwards, there was a concerted effort within social reproduction theory to analyze how unpaid reproductive labor, including care work, domestic work, and sex work, were being reorganized and commodified in relation to late capitalism and neoliberalism. For example, postcolonial scholars expanded upon the insights of the Italian materialist feminists to highlight how intimate and affective labor such as sex work and domestic work is relegated to racialized and feminized populations in the Global South, reproducing capitalist patriarchy on a global scale (Mies 1998;Truong 1990;Parreñas 2011). Prabha Kotiswaran has returned critically to the material feminist tradition to propose a "postcolonial materialist feminist approach" to combat what she calls "sex work exceptionalism"-"the rationale for treating sex work exceptionally by asserting the intrinsic qualities of sexual labor which set it apart even from other types of low-paid, lowstatus work" (Kotiswaran 2011, p. 221). Instead, she proposes a continuum encompassing marriage, sex work, domestic work, unpaid housework, surrogacy, and erotic dancing.
For our purposes, the most valuable insight of the materialist feminist tradition is its emphasis on the economic underpinnings of marriage and its treatment of marriage itself as a type of "work". However, while the model described by the "Wages for Housework" movement has a robust vocabulary for describing how reproductive labor is organized under capitalism, it is less equipped to account for culturally diverse economic systems and kinship relations that do not resemble the nuclear heterosexual family. In the Indian context, sex work is deeply intertwined with complex relations of caste and kinship, which are sustained by endogamous marriage rules. Reproductive labor in general is divided along not only gender but also caste, and heavily mediated by vernacular vocabularies of honor, violence, and patronage. Neither are the social actors in kinship networks fully comprehensible in secular terms: in devadasi 10 communities, for example, it is not just marriage to a man but marriage to the goddess that places women's labor in circulation. 8 For a more sustained elaboration of materialist feminist theorizations of sex work and marriage, see Kotiswaran 2006 9 We use the term "prostitute" here only to replicate Fortunati's own use of it. For our own work, we use the term 'sex work' as this corresponds with the sensibilities and experiences of the people we interviewed and it is a non-stigmatising term chosen by people directly concerned. 10 Devadasi is a pan-Indian term referring to a woman who is ritually married to a goddess, making her sexual labor available to patrons, often of a dominant caste. Devadasi women may be entitled to the privileges of sons after they are dedicated to the goddess.
To engage fully with these realities, we turn to two exchange-based theories of value and labour, namely Gayle Rubin's theorization of marriage as a "traffic in women" (Rubin 1975) and Uma Chakravarti's elaboration of the model of Brahmanical Patriarchy (Chakravarti 1993).
(2) Gayle Rubin and the "Traffic in Women" Gayle Rubin, writing in the context of 1970s second-wave feminist anthropology, offered an influential theorization of marriage as a form of "traffic in women" in a way that contrasts with the materialist feminist models. Her model built partly on gift theory in anthropology, developed in the tradition of Durkheim, Marcel Mauss ([1925] 2002), and Lévi-Strauss ([1949] 1969). Departing from a Marxist understanding of the individual subject, gift theory argues that markers of social personhood, such as gender, are produced relationally, through exchanges of gifts between social groups. "Gifts" in this sense do not refer to disinterested donations, but exchanges of symbolic and material value, which bind givers and receivers in relations of reciprocity over generations. It is not just material value but prestige, personhood, and social capital that are bound up in these exchanges. Levi-Strauss ([1949] 1969 argued that marriage offered the most exemplary instance of gift exchange, since in most societies, kinship, and social structure itself, was reproduced and consolidated by the circulation of women as wives and daughters between men. Rubin revisited Levi-Strauss' model from a queer feminist standpoint, arguing that the category of "woman" was itself produced by marital exchanges, through culture-specific "sex/gender systems" governing the family form. In patriarchal societies both within and outside capitalism, she argued, humans acquire gender identity only through socialization geared towards converting them into "husbands" and "wives" who must produce "boys" and "girls" of their own. Thus, marriage is itself a "traffic in women" that reproduces compulsory heterosexuality and binary gender, and non-heterosexual relationships become a form of social death, excluded from circuits of exchange. Like Fortunati, Rubin identifies "normal" marriage and kinship itself as the source of patriarchal oppression, where gender acquires its material and symbolic value. However, her model (and gift theory more generally) focuses on how value is produced relationally and through social networks, not only labor. We apply Rubin's theory in our analysis by focusing on active kin-making practices, such as marriage, abandonment, adoption, and the building of brothel-based networks, to highlight how kin relations are always political, contested, and in process. To treat kin-making as an ongoing dynamic contested process problematizes the distinction between "real kin" and "fictive kin", pointing out how all kinship is the result of social interventions, not a primordial, static biological relation. While biological families and affinal families might abandon their daughters and wives, women forge various other kin networks in the course of work and migration that contest the primacy of biological and affinal kin.
(3) Brahmanical Patriarchy Rubin's model, located within the milieu of second-wave feminist anthropology, did not consider how the category of "woman" was splintered across class, race, and other axes of identity. In India, to analyze how both gender and caste-based divisions of labor are formalized through kinship, we turn to the model of Brahmanical Patriarchy, elaborated by feminist historian Uma Chakravarti (1993). Chakravarti traces how in the period between 1500 and 1000 BC, following the Aryan invasion, new systems of social stratification emerged whereby Aryan women were distinguished from slaves and persons of lower castes in terms of sexual and domestic labor. She returns to B.R. Ambedkar's emphasis on the role of endogamy in caste oppression (Ambedkar, B.R. [1916] 2013), to show how marriage both circulates women's labor and endows it with different degrees of honor and respectability. In her paper, Chakravarti highlights how Hindu scriptures permit hypergamy, where upper-caste men marry subordinate caste women, but not hypogamy, where upper-caste women marry the subordinate caste. Moving upwards in caste, women accumulate symbolic honor and respectability, and their sexual and domestic labor is withdrawn into the private sphere. Moving downwards in caste through marriage, however, women remain publicly available as sexual and social labor even after marriage, so honor and respectability remain an empty promise. These patterns of exchange result in a division of labor and wealth between castes that subjects members of subordinate castes to humiliating forms of labor meriting unique and violent kinds of force. Since the caste system relies heavily on who women marry, scrutinizing them for signs of dishonor, coded as sexual availability, is key to maintaining caste purity.
Like in Rubin's "traffic in women", the model of Brahmanical Patriarchy as elaborated by Chakravarti, locates caste and patriarchal oppression in marriage and kinship systems. However, the model of Brahmanical patriarchy splinters and complicates the monolithic category of "woman" by showing how it intersects with caste. The model of Brahmanical patriarchy also critically interrogates the notions of "women's honor" and "respectability" of labor, by showing how sexual stigma and humiliating or abject forms of work become symbolically interlinked. Several anti-caste scholars and activists have drawn on the concept of Brahmanical Patriarchy to ask whether sex work can in fact qualify as "dignified labor" in the Indian context. Arya and Rathore (2020) have recently argued that when women of subordinate castes are left only with the option between different kinds of undignified and abject work, individual choice and sexual freedom are "bourgeois abstractions" touted by dominant-caste feminist defenders of sex work (ibid., 4). They also point out the relation between violence in sex work and the dominant-caste violence against Dalit-Bahujan 11 women in the public sphere (Pawar and Moon 2014, cited in Arya and Rathore 2020, p. 4). In a similar vein, Abhinaya Ramesh (2020) has argued that that the legitimation of sex work reproduces structures that extract sexual and reproductive labor from Dalit-Bahujan women and direct it towards dominant-caste families. These scholars rightly question the stagnant dichotomy between "consensual choice" and "coercion" that structure the terms of sex work advocacy, while also challenging researchers to think more deeply about the questions of stigma, respectability, and violence when it comes to work 12 .
While our set of interviews does not have sufficient data to allow us to make general claims about the relationship between caste oppression and sex work, we respond to these challenges by taking our cue from Rubin's and Chakravarti's exhortation to interrogate marriage economies, and from Indian feminists who have long theorized the connections between caste, gender, marriage, and labor 13 . We ask: how did our respondents contest the regimes of respectability associated with sex work by resignifying symbols of marriage? How do neoliberal vocabularies of sex work intersect with vernacular notions of labor and dignity, in India, where 92% of all work (including sex work) occurs in the informal sector (Kotiswaran 2019)?

Applying These Models to the Indian Context
In the past few decades, Indian feminists have grown increasingly polarized on whether sex work should be recognized as work. On one side is the radical feminist approach that conflates all sex work with violence and slavery and admits of no possibility of voluntariness on the part of sex workers. This position has grown stronger with the rise of international abolitionism, evinced by the growing movement within India to criminalize demand and prioritize "rehabilitation and rescue" efforts, led by neo-abolitionist NGOs. It overlaps with more middle-ground feminist perspectives, such as that of the 11 The term dalit refers to those formerly classified as "untouchable" castes and was popularized by the architect of India's constitution and anti-caste theorist B.R. Ambedkar in the 1930s. It was reclaimed in the 1970s as a term of self-identification and political empowerment and mobilization. Dalit-Bahujan refers to the grouping of untouchable castes who refer to themselves as Dalit and to the other backward castes who take on the term Bahujan. 12 For an overview of how debates around caste and sex work have unfolded within Indian feminist groups, see Gopal 2012. 13 For an introduction, see Rege, Devika, Kannabiran et. al. 2013.
National Commission for Women, which acknowledge a distinction between forced and voluntary engagement in sex work, but remain skeptical of the latter, seeing it as driven by poverty. Unlike its Euro-American counterparts, the radical feminist position in India has had a distinctively socialist feminist analysis. For example, early iterations of this perspective, such as the intervention by Jean D' Cunha (1991), argued that sex work epitomized not just patriarchal but economic slavery, and that "choice" could only be a result of poverty. She implicitly invokes a familiar characterization of India's economy as feudal and patronage-based, in contrast to the perceived liberal economic climate of Europe and America, where sex workers are capable of entrepreneurship and choice.
On the other side is the pro-work position that has developed over the past three decades in India, spearheaded by sex workers collectives such as Durbar Mahila Samanwaya Committee (DMSC), National Network of Sex Workers (NNSW), All-India Network of Sex Workers (AINSW), and supported by community-based organizations, NGOs, and sex worker activists. While the pro-work position received its impetus from the HIV/AIDS projects of the early 1990s, it has grown into an international movement with a strong demand for decriminalization and for recognition as workers (Lakkimsetti 2020). Sex worker collectives insist on a distinction between voluntary sex work and coercion or trafficking, advocate for the decriminalization of sex work, highlight their connection to other informal sector workers, condemn police violence, and demand social benefits from the government 14 .
We suggest that focusing on marriage economies and the politics of kinship can challenge the consent-versus-coercion binary that dominates these debates on sex work and trafficking. The theoretical models we use in our paper do not address sex work specifically, but they highlight the economic and extractive underpinnings of marriage, and how it reproduces caste-and gender-based divisions of labor more generally. Each model directs us to ask certain questions about sex work in the Indian context, where almost all cis-female sex workers have been in heterosexual marriages. In the following section of the paper, we identify three sites where the continuities between marriage and sex work become most evident: (a) migration (intra-state, inter-state, and cross-border), often simultaneously motivated by marriage, sex work, and other kinds of work; (b) the regimes of honor, respectability, and stigma attached to women's work, where stigma is often coded as sexual availability; and (c) the multiple kin networks that women sustain or dissolve, as they navigate different labor markets. The three women we describe each had different experiences of migration and marriage. K, introduced in the beginning of this paper, had migrated to a neighboring district after marriage, and wore a mangalsutra to navigate attributions of honor and stigma. M, who had eloped from Bangladesh, managed stigma by concealing her Muslim identity. Estranged from her natal family, she had built her kinship networks in Budhwar Peth. The third woman, J, came from a devadasi community, and identified strongly with her marriage to the goddess Yellamma. For all of them, sex work was not an exception to marriage. The distinction between marriage and sex work was a fraught and contested issue, which they continually managed, navigated, and engaged critically and reflexively.

Methods
The empirical data for this paper were collected by Dr. Sutapa Majumdar, one of the coauthors of this paper, and are drawn from interviews with women doing sex work in Pune, Maharashtra, conducted as part of her doctoral research between January and May 14 For an overview of the Indian debates, see (Kotiswaran 2008), Introduction.
2016 15 Based in feminist economics, her doctoral dissertation aimed to understand the everyday lives of women in unorganized labor markets, with a specific focus on women working in beauty parlors in Pune city, sex workers across the state of Maharashtra, and embroidery workers in Bengal. She approached respondents through Saheli Sangh, a nonprofit working for the rights of sex workers in Pune city. Women were interviewed in the office, as well as in brothels, lodges, tea stalls, and residences, using a snowball sampling technique to recruit participants. Respondents were randomly selected across age, caste, and location of sex work. In-depth interviews were conducted with women who volunteered to share their life stories, using semi-structured questionnaires. Questions focused on their survival strategies within the labor market and how they negotiated with various intermediaries inside and outside their workspaces. She conducted participant observation and focus-group discussions at these settings, in addition to analyzing quantitative data on income and work histories. Verbal consent was obtained for all the research, and narratives were shared with respondents after transcription. The quantitative data were analyzed in MS excel and SPSS, and the qualitative data were analyzed using MS Access.
Of the fifteen sex workers interviewed for the study in Pune city, all had been married. Their migration histories were tied to both marriage and sex work, and the sample included migrants from rural Maharashtra as well as other states within India (Bengal, Bihar, Andhra Pradesh, Karnataka, and Tamil Nadu). Two respondents were from Bangladesh and Nepal, respectively, but had been in Pune for over two decades. Most of them had worked in unorganized labor markets from a very early age, possessed low education, and viewed sex work as a better earning opportunity than other kinds of work they had done in the past.
(1) K K, whose story we began with, came from a poor family in rural Maharashtra and her first migration away from home was to a nearby village, after her marriage. K had the most conventional experience of marriage, being married to a man of her parent's choosing, and from the same community. Her experience also represents the most common kind of migration in India: intra-state migration for the purpose of marriage. According to the 2011 census data, 455 million people in India are migrants, most of them internal migrants. 16 Furthermore, 309 million of India's migrants are women, which means that there are about twice as many women migrants as there are men. Two-thirds of these migrant women (66%, 205 million) are classified as marriage migrants, and the rest as migrating for work-related reasons.
However, K's experience troubles the census distinction between "marriage migrant" and "economic migrant" by highlighting the economic underpinnings of marriage 15 A note on our positionality: Dr. Sutapa Majumdar has a PhD in Economics and many years of experience working in the field of sex worker rights and strongly believes that sex workers too have equal rights to live a life of dignity and respect. Having worked so closely with sex workers outside of academia, she struggled with treating them as "research subjects", straddling the line between objectivity and subjectivity. Dr. Shakthi Nataraj, co-author of this paper, has a Ph.D. in Anthropology, and has worked with sexual and gender minority activists in Tamil Nadu. Her doctoral research was on how thirunangai (transgender women) activists in Tamil Nadu reappropriate and reframe international discourses on transgender rights and sex worker rights. In collaborating with Sutapa on this paper, she brings her own interest in queer kinship and exchange theory to analyze how sexual labor is intertwined not only with adjacent labor markets, but also with marriage and kinship systems. Dr. Nataraj and Dr. Majumdar currently work together on the Laws of Social Reproduction project at King's College, London (see more at https://lawsofsocialreproduction.net/). 16 According to the 2011 census, 62% of India's internal migrants have migrated within the same district, 26% between districts in the same state, and only 12% between states. Only 5.3 million Indian residents are international migrants, comprising no more than 0.4% of the national population (De 2019). The largest proportion of international migrants come from Bangladesh (2.7 million) and Nepal (800,000 persons). more generally 17 . She was married young because her parents could not afford to support her. When she left her abusive husband and returned to her parent's house, she gave birth to a son. This prompted her in-laws to initiate a reconciliation, but the abuse began again when she had a daughter. Thus, acceptance in her husband's family was contingent upon her reproductive labor, specifically bearing a male heir.
Her work history also preceded her marriage and continued alongside it. As a young girl, K, like many other members of her family, was engaged in either agricultural work or domestic work for families of the dominant Patel caste 18 in her village. In addition to meagre payment on a daily wage basis, they subsisted on the grains, bhakri 19 and vegetables they received from their employers. After leaving her husband for the second time, she did domestic work in a Patel household before moving to Delhi and joining sex work with the help of a friend.
Like for most other sex workers, K's engagement in sex work occurred as part of a continuum of livelihood options including agricultural work, domestic work, and peer education 20 .
After five years in sex work, she realized that sex work fetched much more money than agricultural and domestic work:

After five years I came back to the village to work in the fields and as a domestic worker. But the money was not enough, and people in the village already knew what I had been doing, so I came to Pune to work in Budhwar Peth. Initially I worked in a brothel on 50:50 basis and then started taking clients on my own as I got a room to myself. Sex work is just like running a garment shop. It is like a business and people will come to you if you behave well and are able to satisfy your customer. When I was young and had a lot of clients, I used to spend all money on drinking and in gambling. But then after joining Saheli, I realised the value of savings and that helped me transform. Now everyone here in Budhwar Peth loves me and respects me as they have seen how hard I have worked for my children and my family.
In comparing sex work to running a garment shop, K was pointing out the continuities between sex work and other kinds of services. As a peer educator with Saheli, she was also arguing that sex work should be considered work, foregrounding values such as savings, business acumen, and entrepreneurship.
K was navigating a fraught space where she was continually being scanned for signs of "respectability" and sexual availability, coded through signifiers of caste and marriage, attached to labor. Svati Shah has described the pervasiveness of the symbolic linkages between reproductive labor, caste honor, and marriage in her ethnography of construction work markets in Mumbai. Daily economic competition was expressed through speculations about women's chastity, accusing women of specific caste and tribal groups of using 17 As Ravinder Kaur (2012) cautions, a watertight separation between marriage and labour means that migrant women are consigned to being studied only as the "trailing wives" of male migrant workers, rather than as active agents of migration themselves 18 Patel connotes a caste community of land-owning farmers, businessman, agriculturalists, and merchants. Members of the Patel caste reside primarily in Bihar, Maharashtra, Madhya Pradesh, and Delhi. Traditionally the surname bestows status, referring to the village chieftains. 19 Flat bread made of millets 20 We use the term "continuum" following anthropologistSvati Shah, who has long argued that sexual commerce in India exists as part of a continuum of informal labour practices and livelihood strategies pursued by poor migrants (2014, p. 113, see also Shah 2007Shah , 2008. her ethnography of sex work in Mumbai (2014) she discusses how brothel-and street-based sex work occur alongside construction work in the context of gentrification and rural-urban migration driven by agrarian distress.The Pan India survey of 3000 sex workers offers powerful quantitative evidence that most women in sex work have a long history of working in other labour markets, and often pursue sex work alongside other livelihood options including domestic work, petty vending, and other informal sector labour (Sahni and Shankar 2013, p. 45). sex to undercut wages for all. Shah writes that "while only a fraction of women solicited clients for sexual services… women understood that by being poor, not upper-caste, and sitting on the street to solicit work, they were apprehended as being potentially available for sex" (Shah 2014, p. 76, italics ours). K was all too aware of such perceptions. By wearing the mangalsutra, she actively pre-empted the stigma of her work, while still fulfilling her obligations to children and family.
She had a critical and reflexive relationship with her "wife" identity, being all too aware that the safety provided by marriage was always contingent upon her labor, and the "honor" attached to it. K had broken kin relations with her husband's family, and instead built networks with her natal family and fellow sex workers at Saheli. At Saheli, she had earned respect not as a "wife", but by fulfilling her obligation to her children and natal family, by getting them educated and married.
(2) J The second woman whose story we want to consider, J, was from the Matang 21 caste and came from Belgaum, in Karnataka. She hailed from the jogati 22 community in Northern Karnataka and had been married to the goddess Yellamma soon after puberty. J belonged to the jogati community similar to the one studied by Lucinda Ramberg (2014), where families have the option of dedicating their daughter to the goddess Yellamma rather than marrying her to a man. Following their marriage to the goddess, they lose the respectability afforded to housewives, cannot marry men, and become available for sexual non-marital relationships with upper-caste patrons. They also acquire privileges and responsibilities usually restricted to sons, such as inheriting land, retaining their children, and supporting their families. Ramberg uses a theoretical framework based on Rubin's "Traffic in Women" to argue that marriage to the goddess forms a parallel sexual economy to that inhabited by housewives, enabling landless Dalit-Bahujan families to retain wealth and land within the natal family.
Because of her jogati background, the continuities between sex work and marriage took a very different form for J than for the other women interviewed. Her marriage did not demand migration to a husband's house; she continued to live at home, doing daily wage labor alongside her mother, brother, and uncles. A few years after her marriage, she began a relationship with an unmarried man and had two children with him, but as a jogati or devadasi, she could not marry him. Instead she moved to Pune with her children.
Since I am a devadasi, I could never marry him. He was in love with me and didn't want to get married to anyone else. I didn't want that and so came to Pune with my children. I wanted him to marry and start a family. I am a devadasi. Whoever gives me more money, I will go with him. I can't stay with one man forever. That's what we devadasis do.
J's rejection of her male lover, and her subsequent migration to Pune, was a sign of commitment to her marriage to Yellamma. J saw herself not as a woman who had been injured, deceived, or wronged, but as a person performing the ethical action demanded by their kinship role. In saying "that's what we devadasis do," she was contrasting the 21 The terms Matang and Mang are used synonymously to refer to a Dalit caste, the members of which mainly reside in the Indian states of Maharashtra, Andhra Pradesh, and Karnataka. Mangs were identified as "criminal castes" by the British colonial Government, in addition to facing widespread persecution by upper-caste Hindus. In 1952, five years after India's independence from colonial rule, the Mang caste was decriminalized, and in 1961 they were listed as a "scheduled caste". 22 Jogati is a term for a person ritually married to the goddess Yellamma. Yellamma is a popular Hindu deity in in the southern states of Karnataka, Andhra Pradesh, Tamil Nadu, and Maharashtra, associated with jogati and devadasi communities.
immobility imposed upon women married to a man, to the freedom of movement afforded to a devadasi, while highlighting parallels between the two marriages. 23 Like other women in our study, J also had to manage symbols of kinship, honor, and respectability. While being a jogati meant she would never command the respectability of a housewife, it also meant that her moment of entry into sex work was not seen as a loss of symbolic honor in the same way. In the case of K, for example, who was married to a man, her mangalsutra was symbolically at odds with her doing sex work. J, on the other hand, saw sex work in alignment with the sexual economy and kinship role she had entered. She continued to wear her moti 24 , which communicated that she was a devadasi.
When K, the first woman whose story we presented, left her husband, she took her children with her despite his family's wishes. In J's case, however, the father of her children never had a claim upon them. Her claim over her children highlights the fact that her marriage to the goddess was just as "real", in the sense of being materially consequential, as marriage to a man. Both forms of marriage placed women's reproductive labor into circulation in specific sexual economies, and both produced certain possibilities for kinmaking practices.
In J's case, she could not build kinship networks with the father of her children. However, in Budhwar Peth, she had met many devadasis from her village, and she fortunately had a relatively supportive brothel-keeper who gave her a room to stay, brought clothes and food for her children, and took care of the children when she did sex work. She had built kinship networks within Budhwar Peth over the years, and continued to have a cordial relationship with the father of her children, as well as with his wife and family.
While Yellamma women tend to be portrayed by reformists as victims of superstition, backwardness, poverty, and caste violence, J viewed her marriage to the goddess as an indicator of her value and power 25 . She said, half-jokingly: Being married to Goddess Yellamma and being a devadasi separates me from the rest of the women in dhanda (sex work). My clients are mostly good, as they fear that being a devadasi, I could bring bad luck to them if they don't keep me happy. I have special power you see!(laughs).
Still, J's relationship with her devadasi role was not without conflict. She got her daughter married at a very early age, as she did not want her to be either a devadasi or a sex worker. While she no longer did sex work, at the time of the interview, she worked as a peer educator at Saheli and earned a fixed salary. Like most other women, she combined this with other informal sector work such as making and selling food, and unpaid work such as caring for her grandchildren.
J's experience was certainly quite different from that of other women interviewed; it nevertheless challenges sex work exceptionalism and directs attention back to marriage as an economic institution. In her study, Ramberg argues that jogatis do not represent a special case of kinship-based oppression, but that kinship and caste are the source of violence and extraction for both Yellamma women and housewives. Scholars studying other communities where forms of sex work are sanctioned through marriage and kinship structures have similarly argued that any critique of such practices must be based in a critique of patriarchal kinship and marriage more generally (Agrawal 2008). This point is important to make, because in recent years, feminists concerned with the implementation of the Sustainable Development Goals have problematically described kinship-based sex 23 J's comment about how migration is part of being a devadasi recalls an anecdote shared by Durgabai, one of the Yellamma women Lucinda Ramberg describes in her ethnography. Durgabai compares her own freedom to roam the fields like an "untethered bull", to the restrictions placed on women married to men, who needed permission to "move an inch" from their house (Ramberg 2014, p. 154). 24 A beaded necklace signifying devadasi identity and marriage to the goddess. 25 J's experience aligns with Ramberg's argument that many Yellamma women have models of ethical aspiration that pose a queer challenge to secular liberal conceptions of what "real" marriage and progressive politics should be.
work practices, such as that of the Bedias, as "child sex trafficking" (Dalla et al. 2020). Using simplistic binary of choice and coercion, the authors identify parents as perpetrators of sexual exploitation, and argue that anti-trafficking NGOs should eliminate this practice. Attention to the economic underpinnings of marriage can combat this problematic analysis by focusing on structural power imbalances grounded in intersectional gender/caste dynamics.
(3) M The third woman in our study, M, was originally from rural Bangladesh. M was from a Muslim background and grew up working in the paddy fields of wealthier families. Moving to Dhaka to work in a garment factory, she met an Indian man who transported goods across the border and eloped with him to India. They got married in a local temple, and she began living as a Hindu married woman.
While K's migration experience was common, M belonged to a very small minority. Only 0.4% of the national population is comprised of international migrants (De 2019), and M had long contended with the xenophobic and Islamophobic discourses levelled at Bangladeshi women in Pune, so she was initially reluctant to share her migration history.
Marriage was M's entry point into both migration and sex work. It is important to note that M saw her own experience not as an exceptional instance of trafficking but rather as a betrayal by her husband. Her experience of force, deception, and being "sold", should not be read as an exception to "real" marriage, but as a possibility within the structural logic of marriage itself. M's experience is not uncommon. In her study of migrant sex workers in West Bengal, Guha found that many had been coerced into sex work by members of their household and communities (Guha 2018b, p. 58). Guha argues that the violence in sex work was only one phase of a "life cycle of violence": "a spectrum of experiences of everyday struggles with power inequalities and violence, which preceded and persisted beyond sites of sex work across family, state and market situations", beginning with "daughter disfavour" (Guha 2018b, p. 57). Most women who are "rescued" by anti-trafficking interventions find that repatriation to their home communities or families involves more violence than sex work (Ray 2018). M's experience also highlights the contingency of the safety and honor provided by marriage, experienced not only by sex workers, but also by other women who are on the margins of marriage economies. In state rescue homes, for example, sex workers are clubbed together with women who have lost their "honor" and marriageability, falling into other parallel sexual economies. These include minor girls who had eloped against their parents' wishes, rape survivors, divorcees, and persons with disabilities (K. Sudha 2016; Guha 2018b).
Far from longing for the domesticity of a married woman, over the years, M had come to see marriage itself without sentimentality, as an instrumental institution. She highlighted the overlaps between marriage and sex work, when it came to violence, deception, and the extraction of women's labor. 26 She phrased it succinctly, saying: "I realized very early in life that marriage is just an eyewash. It's not difficult to lure poor women into marriage and then make them work to earn money." As in the case of K, M saw her marriage as an "eyewash", but she drew selectively on symbols of marriage and religion to help her work. For example, as she migrated between cities doing sex work, working as amehendi 27 artist, waitressing in a bar, and doing domestic work, she presented her religion differently based on context. Contesting the primacy given to marriage and religion when it comes to women's identities, she ironically referred to work as her primary religion: Those who come to do sex work, for them work is their only religion. If being a Hindu, I get more money, I will become a Hindu. If being in veil makes me sensuous and desirable to my clients, I will be a Muslim. If by chance I get to service a gora (white/foreign national), I will say hail Jesus! M's ironic statement highlights the way that adopting religious markers (such as a veil or mangalsutra) can help manage the regimes of honor and stigma structuring women's labor. The irony derives from the fact that for M, adopting such markers is not simply a strategic masquerade. After all, M had indeed been married to a Hindu man and lived as a Hindu wife before being sold to a brothel. M is pointing out that the "eyewash" is located not in her shifting presentation of self, but within "real" marriage itself, which masquerades as a site of romance, religious sanctity, and respectability, but is in fact an economic institution, and a site of work. With her playful exclamations ("hail Jesus!") M contests the meanings and value attributed to symbols of marriage, religion, and respectability (such as the veil), exposing them as tools of work. With her mocking tone, she distances herself from the value system of her clients, showing that while they prefer to retain the fantasies of sleeping with a religious woman, "work is [her] only true religion." Anti-trafficking interventions would see M's break from her natal family and husband as an unmitigated loss of kinship, to be remedied by "rescue" and repatriation back home. This mistaken understanding inadvertently reinforces the hegemony of the heterosexual conjugal family, and reproduces a colonial logic criminalizing other forms of patronage-based kinship (Freitag 1991; Hinchy 2014) Over the years, M had quite consciously disinvested in relationships with biological family and male partners, investing instead in brothel-based kin networks. Despite her violent introduction to brothel life, eventually her brothel-keeper became something of an ally, allowing her to wait until she was ready to start sex work, and helping her to save money and earn quite well.
M's experience resonates with historian Ashwini Tambe's point that in India, kinship-based idioms in brothels are not an imitation of "real" families but arise from actual material and historical continuities between women's families and brothels. Tambe's study shows how in Mumbai in the 1920s, the brothel became a residence for women thrown out of home, as well as their regular clients, children, and other workers, functioning as a family unit (Tambe 2006). With the support of her brothel-based kin, M's children were educated, and two of them were married. She no longer did sex work but worked as a peer educator at Saheli, fostering kin networks with other sex workers and activists. 26 Nalini Jameela, author, activist, and sex worker, expresses this eloquently: "Getting married isn't any guarantee against violence, but there is a general feeling that one can somehow bear violence from a husband, but not violence from a client" (Jameela 2007, pp. 139-40). 27 A form of body art where intricate designs are drawn on the hands and legs with a paste made from the leaves of the henna plant. Mehendi designs are not permanent, and only last a week or two. They are associated with festivals and auspicious occasions like marriage and childbirth.
M's experience contradicts the assumption made by some researchers that women in sex work aspire always to marital domesticity and remain unmarriageable against their will. 28 M had intermittent relationships and proposals for marriage, but her frequent migrations for work made these hard to sustain. In Pune she had two children with a man, but did not want to formalize her relationship as marriage. Like many other women in sex work, M had negotiated arrangements that allowed her to earn independently and remain mobile, while engaging in a heterosexual partnership on her own terms.

Conclusions
In this paper, we approach the debates on sex work within India at present by shifting attention away from sex work, narrowly conceptualized, to the economic underpinnings of marriage as a form of unwaged labor circulated between families.
Examining the marriage economy as a way to study sex work invites research that complicates the conventional narrative of the abject cisgender woman pushed into sex work through sheer poverty, admitting of more diverse subjectivities and locations from where people exchange sex for money, whether middle-class women in Bangalore wanting to send their children to English-medium schools (Vijayakumar,orthcoming), or those outside the ambit of sex worker activist networks. Sex work is also prevalent amongst sexual and gender minority communities, who are powerfully implicated in circuits of marriage, kinship and exchange, albeit in ways different from cisgender women. These include kothi 29 and transgender sex workers who run away from home to escape marriage pressure or who support their biological families with sex work earnings, kothis married to cisgender women, or non-biological kinship forms like the joint family system amongst transgender women in India. A focus on marriage economies and sex work moves beyond narrow questions of sexual identity, to encompass migration, labour, and a broader consideration of queerness as social marginality.
Our paper directs attention also to how marriage-related regulation regimes are emerging alongside other transnational abolitionist anti-trafficking regimes. The 2000 UN Protocol to Prevent, Suppress, and Punish Trafficking in Persons, Especially Women and Children has given rise in the past two decades to a transnational field of anti-trafficking law led by U.S. interests, aimed at abolishing sex markets worldwide and advocating carceral approaches to cross-border trafficking. This global edifice has come under critique for its disproportionate focus on the exploitation of sexual labor, rather than other kinds of labor (Shamir 2012;Kamal Prasad 2018); its conflation of voluntary migration and sex work with trafficking for the purpose of sexual exploitation (Nayak 2018; Bandana Pattanaik and Sullivan 2018; Pai, Seshu and Murthy 2018); and its carceral approach (Kotiswaran 2019). Scholars studying migration and sex work have argued that in the guise of "sexual humanitarianism" (Mai 2018), anti-trafficking regimes enable states to enact violent anti-migration policies, creating artificial divides between wives, sex workers, and domestic workers that treat them as victims rather than as workers (ñHwang and Parreñas 2018).
The ILO and UN's 2017 inclusion of "forced marriage" under the sign of "modern slavery and trafficking (ILO & Walk Free Foundation 2017), might signal a worrying expansion of powerful governance feminist instruments to manage kinship and family relations. As mentioned above, newer research in India has begun to pathologize Bedia and other communities under the rubric of "child sex trafficking" (Dalla et al. 2020), and the characterize the rise in bride migration to Punjab, U.P., and Haryana as "sexual slavery" or "bride-buying", conflating child marriage with trafficking (cited in John 2020). We join other Indian feminists in critiquing these trends, pointing out the economic drivers of 28 For example, see (Ghosh 2017, p. 197). 29 A pan-Indian term referring to a person assigned male at birth but who identifies as feminine to some degree, and who is attracted to men. The term is often used as an "indigenous" translation of the public health term, "Men who have Sex with Men." "normal" marriage itself (Agrawal 2008;Kaur qtd. in Bertram 2015;John 2020). Our research data and their analysis show that we need to continue to critically interrogate how a marriage/sex work binary is deployed to facilitate the migration of some and immobilize others in precarious kinds of labor, particularly in India, where global anti-trafficking discourses have had paradoxical and unique trajectories.

Author Contributions:
The narratives presented in the paper were originally collected as part of the doctoral work of S.M. Both S.N. and S.M. have contributed to the conceptualization, theorization, interpretation and analysis of the data in the present paper. All authors have read and agreed to the published version of the manuscript.

Funding
The present paper was written as part of a project that has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No 772946). Dr. Sutapa Majumdar's doctoral research was made possible with the financial support of the Indian Council Of Social Science Research (ICSSR), New Delhi.

Institutional Review Board Statement:
The narratives analysed in the paper were collected as part of doctoral work conducted at Savitribai Phule Pune University, Pune, Maharashtra. The research was approved by the university research ethics board.
Informed Consent Statement: Verbal informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from Dr. Sutapa Majumdar. The data are not publicly available to protect the confidentiality of the persons involved.