‘Left-behind’ women of migrant-men: Rethinking agency and autonomy of women in rural Bangladesh

This paper details the changes in the agency and autonomy of left-behind women of international migrant-men exploring continuity and changes in the discourse and systems of traditional gender roles of a village in rural Bangladesh. It sheds light on the pattern of trans-border communication of the women and their mobility in the locality and subsequent changes in gender relations. Through ethnographic fieldwork, this paper examined how women rearrange their culturally constructed boundaries and meanings of their gender. This study also explored whether these changes are temporal for the migration period and what happens when husbands return permanently. The study followed the post-structural and intersectional feminist approaches and suggested that men’s migration creates a situation where some women get the advantages of widening their room for manoeuvre, which may be sustained after the return of their spouse.


INTRODUCTION
The migration of men leaving their women is an increasingly common migration pattern in South Asia and other developing countries (Desai andBanerji, 2008, Gardner, 2009). The scholars of international migration studies have increasingly focused on gender issues over the past few decades (Lutz, 2002;Morokvasic, 2004;Shah, 2004). In the beginning, the focus was on female migration and its impact on the sending communities. Bulk of the literature discussed female migrants, ignoring the left-behind women of migrant-men (Hugo, 2000;De Haas and Rooij, 2010). However, the consequence of migration is also experienced very acutely by the leftbehind women of migrant-men (Hugo, 2000;Gardner, 1995Gardner, , 2006. This paper aims to address how women conform, negotiate or resist dominant discourses and practices of gender relations, and express their agency and autonomy in a Muslim society following their husbands' migration abroad. In Bangladesh, migration is often used as an economic survival strategy, especially when life-course events have become unviable. Its members have to move to seek alternative livelihoods, either permanently or periodically (Gardner, 2009). The migration of men gives women a chance to play the role of their absent husbands (Gardner, 1995). This shift may enable these women to become more authoritative, positively influencing their property ownership, productive decision-making, household expenses, and personal freedom (Fakir and Abedin, 2020). While the previous discussions of gender studies focused on issues such as autonomy, empowerment, and emancipation in general, the outcome of recent studies have come to increasingly highlight individualism, multiplicity, and contradiction in the experiences of men and women (Rashid, 2013). Men as remittance providers and women as remittance managers cease to perform in many cases soon after the returns of the migrants, something which is also evident from other contexts (Mckay, 2015). These practices indicate the post-structural ideas to investigate the diversity and contradictions at the individual level. However, Bilge (2010;23) argues that post-structural perspectives do not adequately pose individuals' agency within particular contexts as they avoid taking agency as a self-explanatory concept tied to a pre-discursive self. Consequently, the socio-historical processes of subjectification remain largely underspecified. Therefore, the proposed study has followed both post-structural and intersectional feminist approaches to identify the social and historical context of women"s "agency" and "autonomy," keeping the mobility level right during migration and maintaining the day-to-day life accordingly.

Left-behind wives of migrant-men in existing literature
De Haas and Rooij (2010), in their study in a rural area of Southern Morocco, show that the well-being of left-behind women depends on the economic success of male migrants and the degree of interference from the members of husbands" extended family. Their discussion suggests that although remittances give women more decision-making power, this is mainly temporary as the migrants take over their "patriarch" as soon as they return. They argue that women feel more secure when they live with their husbands because the loneliness, seclusion, long absence of sexual relations, and over-dependence on remittance complicate their lives. The improvement that we see in their position, although accelerated by migration, is principally the result of general social and cultural change (p57-59). Likewise, Rashid (2013), in her ethnographic study in two gulf migrant villages in Cumilla, Bangladesh, argues that women do not necessarily enjoy their decision-making power, autonomy, and freedom in the absence of their spouses. Instead, they like to follow social norms and stay in male protection. Similarly, Seigmann (2010) study in Northwest Pakistan shows that women do not enjoy material rewards because their physical and mental health suffer in the absence of their husbands.
In contrast to the above discussions, Maharjan et al. (2012) in their study in rural Nepal show that migration of men enhances the involvement of women in the society through their involvement in household tasks, bargaining, and buying quality products, saving money and buying movable and immovable assets and investing in wise things that gradually empower them. Similarly, Gardner (1995,2009), in her ethnographic study in Sylhet, Bangladesh, argues that migration of men and financial improvement of the household upgrade women as they get chances to negotiate in the household decisionmaking process. However, her study does not show what happens to these practices when their men return home.
We understood many key issues of transformative impacts of temporary migration of men on the left-behind women for their economic security, self-esteem, and social status through these kinds of literature. However, by focusing on security and empowerment discourses in general, their agenda for analysing women"s experience hides women"s agency and autonomy at the individual level. Therefore, they are not sufficient to understand the variety of women"s preferred way of life and their negotiations with or resistance against patriarchal norms and ideologies while their husbands are abroad (Rashid, 2013).
Practically, the impact of temporary male migration on their women is diversified and hence difficult to understand by one single discourse (McNay, 2016). Economic remittance contributes to reshaping the staybehind women's financial condition and their households (Gardner, 1995). Male migration brings the sender view from abroad that positively changes women's position in the traditional community. This result in women"s involvement in the outside job, enhanced mobility, access to a better health facility, higher participation in reproductive decisions, and reduced domestic violence (Hadi, 2001). Their negotiations in household decisions and interactions in the public sphere create scope for both resistance and alternative discourse of morality and identity in their favour (Dannecker, 2002, 202). A study conducted by Giri and Darnhofer (2010) shows that the women as left-behinds of out migrant-men have shown that they acquire more opportunities than non-left behind women.
This article examines left-behind women's experiences to explore the changes in their agency and autonomy and the sustainability of the practices after the migration period. My argument relies on a series of interrelated concepts: "identity, "self," "agency," and "autonomy" to understand the pattern of agency and autonomy in Bangladesh and South Asia. Sökefeld (1999) shows that identity and self are differently perceived in the traditional anthropological discourse, where the Western self is regarded as autonomous and egocentric, while the non-Western self is shared with others. Intersectionalism emerged as a particular approach in feminist theory to analyse the multiple sources of origin of women"s suppression (Nash, 2008). Sökefeld (1999: 417-18) shows that the previous understanding of non-Western notions does not indicate individual features for which the recent anthropological discourse diverts attention following the post-structuralist approach. However, according to Cohen (1994), emphasizing the self's discussion does not mean indulging in individualism but exploring how the self is linked to groups, society, community, or the shared identity. In this regard, Brah (1996) shows that separate identities and shared identities perform in the same field, staying connected. They are constantly changing and hence are different to different individuals at different times. There is intersectionality between different identities; no one identity is absolutely "pure," but one identity can overwhelm or make another less important. However, each identity will reflect as a group identity when involving agency participation.

Agency and autonomy
Autonomy can be defined as the capacity to manipulate one"s personal and social environment and the ability to obtain information and to use it for their personal and intimate concerns (Dyson and Moore, 1983: 45). Herr (2018: 197-78) considers that an autonomous person is an individual who can set goals for her/his life from a list of feasible options according to will and without external interference. On the other hand, the agency is "the power within" or the ability to define one"s goals and act upon them (Kabeer, 1999). Ortner (2006) also conceptualizes agency with the notion of power to analyse domination, resistance, and inequality. She argues that social actors are neither thoroughly determined beings nor free agents who act outside the social systems. Instead, they are individuals who are not only bound but also enabled by social orders to produce and reproduce social formations of power. Similarly, Tenhunen (2006: 129-130) argues that women"s agency is not obtained from complete denial of, or escape from, cultural meaning. This indicates that culture not only restricts women's agency but also enables it.

Inter-connectivity of agency and autonomy
McNay (2016) focused on dynamism and multiplicity, while discussing the agency of individuals. She argues that individuals' agency and subjectivity are dynamic, multifaceted, fluid, yet unified in daily life. Based on the above discussion, I consider that the idea "agency" and "autonomy" are interconnected. The agency emphasizes the capacity of actions motivated by free will or social conditioning, while autonomy emphasizes reflexivity and self-governance. The study will not focus on the differentiation between agency and autonomy; instead, it will highlight how cultural practices enable women within codes and how they are conscious about their position while dealing with others in the household and the public sphere.

International migration from Bangladesh
The partition of India in 1947 and the independence of Bangladesh in 1971 resulted in the displacement and Uddin 3 resettlement of millions of Hindus and Muslims causing a profound transformation in the region (Alexander et al 2015). One-third of this population took shelter across new borders of the Bengal delta. The other one-third was internally displaced while the rest migrated to the Middle East, Europe or North America (ibid). Therefore, the history of Bangladesh is a history of migration. According to a World Bank report in April 2019, Bangladesh has been recognized as the ninth largest remittance recipient globally and the third in South Asia as it received USD15.50 billion in 2018. The present contribution of remittance to national GDP is 5.4% (World Bank Group, April, 2019). The service sector's contribution to GDP has increased from one-third to two-thirds. Industrial production has increased from 7 to 15%, while agriculture's contribution to GDP has decreased from one-half to one-sixth over the last few decades (Lewis, 2011). The changes in these sectors deeply influenced the lives of the people of Bangladesh. Consequently, internal and international migration and mobility have increased in the country. Between 1975 and 2009, more than 3% of the total population migrated from rural to urban areas every year. This is one of the highest rates of urbanization in the world (Alexander et al., 2015, P. 2 (Sikder, 2008;Bal, 2013).

METHODOLOGY
Together with a research assistant, I conducted ethnographic fieldwork intermittently from January to December 2017 in Rashidpur (pseudonym), a migrant village in rural Bangladesh. We did the fieldwork applying participant observation as the main method of research with other anthropological techniques such as census, Key Informant Interview (KII), in-depth interview, life history interview and informal group discussion through our residence to get a complete idea about changing gender roles in the village. Initially, we conducted a census using a structured questionnaire to gather basic informationage, sex, education, marital status, occupation, monthly income, land and other asset holding, housing condition, residence and duration of migration-of all the villagers including migrant, nonmigrant and aspirant migrant households. It was necessary for an overview that helped us assess the fundamental changes in the migrant households of the village. Subsequently, we collected data and information through continuous observation and interaction, life history interviews, individual interviews, and group discussion throughout the period. Participant observation and "hang out" helped us gain insights into peer relation and various aspects of village culture while life history interview was important to get personal experiences of the informants and individual interview was useful to avoid ideal answers, masked subtleties out of fear, shame or misuse of information (Berckmoes, 2013). Group discussions provided us with a general overview of different issues.
In the study, we investigated how the migrants" wives manage households in the absence of their husbands, how they interact with the public sphere, and how their interaction with men is guided and monitored by the norms and values of patriarchy. In this case, we attempted to systematically observe and participate in all aspects of their daily life events through normal patterns of interaction to internalize their everyday happenings from their point of view, their relation to life and vision of the world (Malinowski, 1922, p25). Particularly, we attempted to observe and participate in the festivals, ceremonies, and other events to understand the gender interaction of the villagers. We also observe a few village arbitrations to understand how the villagers deal with women's issues. We conducted interviews in the form of discussion at leisure time or convenient time of the villagers. In some cases, we interviewed the same person several times, depending on the situation and the depth of the information.
During the fieldwork, we talked to men and women of aspirant migrant, migrant and returned migrant households to assess their transborder communication, gender-based spatial mobility, and the changes in the agency and autonomy of women. We also talked to the village leaders, religious leaders, entrepreneurs, primary school teachers, service holders, drivers, masons, rickshaw pullers, agricultural labourers, and the like to get a complete picture of the gender-based social change of the village. We talked to the villagers face to face sitting in their house, village mosque, village bazar, tea stalls, grocery stores individually or in groups. The entire fieldwork consisted of approximately 200 discussions and group discussions ranging from half an hour to several hours, depending on the situation.

Ethical issues
We tried to maintain the ethical issues of the field. To do so, we did not use camera without the permission of the villagers to show respect for their privacy. In addition, we did not record our conversations during the interviews and discussions because we found that people speak ideally and become formal in front of the recorder. We used only field diaries to note down on important information during the interviews and wrote observation notes on "social spaces." Before talking to the informants, we clearly stated the purpose of the study and sought their consent to collect information with a trustworthy rapport. The discussion was informal and open-ended. After the fieldwork, the scattered interviews and observation notes were read several times and coded according to the purpose of this research. After that, the author expanded the codes and categorized them for drafting the report according to the central theme of the study. Absolute anonymity is significant to guarantee the well-being of the informants (Wilson, 2019). Therefore, pseudonyms of all informants and the village were used in the field notes and the publications when referring to people and places.

The village
Rashidpur is a village under Rashunia Union Parishad in Sirajdikhan Upazila of Munshiganj district, Bangladesh. It is one hour away by bus to the south from the capital city Dhaka. The physical setting of Rashidpur is different from the surrounding villages. The village is situated in a low land area where communication to and from is very difficult as it does not have road connection with the adjacent villages from three directions. A cement road reached the south side of the village. This is the main path of the villagers to communicate with the surrounding villages and outside. Again, there are very few or no built road for communication within the village; there are only unpaved paths to go from one homestead to another. As a result, during the rainy season, the homesteads become separated from each other like islands scattered in water. Therefore, boats are the only means of their communication during monsoon (Uddin, 2018). Since the village is located in a low-lying area and it sinks in the rainy season every year, working as agricultural labourers, cultivating paddy and jute in the land and fishing in the water has been the main occupation of the villagers for many generations. However, the cost of cultivation has become more expensive than the price of crops for which many villagers have lost interest in cultivation. Moreover, as they did not have enough land and their education was low, they were looking for an alternative way of survival. However, due to corruption and nepotism in the recruitment process, it is difficult for these low income people to manage good jobs or run a profitable business in Bangladesh. So, migration has been considered as their best option. Thus, they started going abroad since the early eighties (Uddin, 2018). On the other hand, Mahama (2013) considers excessive poverty, unemployment, natural disasters like severe cyclone, riverbank erosion, poor crop yield, religious conflicts (minority conflict), and political instability as the major factors that caused migration in the 1980s.
Nowadays, migration has become the main source of income for many villagers in Rashidpur. There are many households which are so dependent on remittances that they are unable to manage three square meals without it. Remittances have also transformed the physical structure of the village from thatched to corrugated tin or brick buildings. According to census data, the village has 1404 people who are divided into 307 households, of which 110 are the migrant households. There are 133 (127 men and 6 women) migrants who have migrated to different countries, particularly to the Middle East, Southeast, and East Asian countries. The village also has 55 returned migrants in 48 households. Most of the villagers are Muslims. There are only six Hindu households, four of which have migrant-men (Uddin, 2018). This study was conducted in two categories of households" eklodge or extended household where women live with their in-laws and bhinno or nuclear households where the left-behind wives of the migrant-men live alone with their children. This article will briefly outline the experiences of women from both categories of the households to understand the continuity and changes in the discourse and practices of traditional gender roles and gender relations in rural Bangladesh.

Agency and autonomy of the wives of Ekloge (extended) households
Marriage brings fundamental changes in women's lives in South Asia, as they are transferred from their natal homes to that of their in-laws (Rashid, 2013). After marriage, they realize that they are in new places where everybody is unknown, sometimes including their husbands. For this reason, they frequently visit their natal homes and want to stay there more time until they become mothers. When they become mothers, slowly, they begin to think of their husbands" households as their own. The ethnographic data of Rashidpur village show that the wives who live in extended households are usually young whilst, in most cases, the wives of nuclear households have adolescent or grown-up children. Remittance of ekloge households is received by the father or brother or, in some exceptional cases, by the mother or sister of the migrants. Decisions are taken by mutual sharing between the recipients, older household members, and the migrants over the phone from abroad. New wives do not actively take part in household decisions. Sometimes, they cannot take independent decisions, even on their issues. In an individual discussion with Laila (research assistant), Taslima (23), a young wife of an ekloge household, shared her feelings that her husband, who lives in Kuwait, sends remittance to her father-in-law. She is aware that her husband sends the remittance and informs her in-laws over the phone. However, the husband does not feel the need to inform her although they have passed five years after marriage. The case represents the situation of the majority of young wives of ekloge households where women are not involved in the household decision-making process and dealing with money. The young women are mainly responsible for taking care of their children and elderly sick members and performing household duties supervised by their mothers-inlaw or other older male and female members of the households.
Young wives of ekloge and rich households visit markets with household men during Eid or other big festivals to buy clothes, shoes, ornaments, or cosmetics as people consider these as occasional and higher status. However, they are not seen to go to the daily bazaar to buy necessary commodities from grocers or fish sellers, or vegetable vendors. Women"s bargaining with these men is undermined by the norms and values of their household men. One afternoon, Karima (22), whose husband (35) has been living in Saudi Arabia for 15 years, told Laila that her husband"s household is a traditional one where the members like to live in an extended arrangement for generations. Her father-in-law has four brothers. The brothers got divided from extended arrangements eight years ago when the household turned The findings of the study reveal that in some cases financially rich and traditionally extended household men treat their migrant son"s young daughters-in-law very politely. The men in this upper-class society take their women out and get them rich dresses, ornamentals, quality food items, etc. They expect their women to go outside the home wearing burqa to maintain pardah. They consider it gives more protection and security to their young women. The women also respect the feelings of the household men in many cases and behave accordingly to maintain the household tradition. In another private conversation with Laila, a wife in an ekloge household revealed that she had lost freedom of mobility after marriage. She enjoyed the freedom of moving around thoroughly when she was in school. After marriage, her movement got restricted. In this regard, Gardner (2009) argues that women from affluent households have to strictly follow the rules of patriarchal society. For many wealthy men, women"s mobility out the homestead is not regarded as higher status for them. Although the women of these households are willing to go outside, they do not do so violating their household tradition. Rashid (2013) revealed that women's power and agency should be evaluated within the discourses or structures of gender relations in rural Bangladeshi society, which expects women to be "good daughters-in-law," "good mothers" or "good wives. She mentions that the "good daughters-in-law" of extended households are judged by their "modesty," "responsibility" and "care attitude" to their in-laws. On the other hand, women of nuclear households are judged by their efficiency, intelligence, and ability to manage them. However, her study shows that women are physically safe, socially cared and economically secured when they live in extended households. Therefore, she suggests that women like to follow social norms and stay in male protection instead of enjoying decision making power, autonomy, and freedom in the absence of men (Mahmood, 2005;De Haas and Rooij, 2010;Sinha et al., 2012). McNay (2016) argues that by overemphasizing religion in Muslim women's lives, the scholars tended to minimize other forms of agency in investigating multiple social structures, expectations, and pressures in their daily lives. Therefore, she suggests a dynamic, multifaceted, yet unified subjectivity to investigate more realistic formulation of agency. McNay's theoretical ideas are applicable in the ethnographic accounts of the women in Rashidpur village. Close observation of the activity and mobility of the members of migrant households and their intra and inter-household relations shows that the impact of male migration is diversified on the members of migrant households and their living arrangements. This idea indicates conflicts, arguments, and counter-arguments in the household and society that Rashid, Mahmood, and De Haas and Rooij overlooked. In this regard, Akram and Karim (2005) show that although male migration ensures women's economic security and social status, it contributes to increased control by their in-laws. Therefore, although there are good relations in some exceptional cases, in most cases, interpersonal conflict arises between daughters-inlaw and other members, which ultimately contributes to the breakup of the households after migration.

Agency and autonomy of the wives of Bhinno (nuclear) households
The women's experience of their husband"s migration varies from household to household depending on their age, educational background, personality, amount of remittance they receive, and above all, their relationships with their husbands (White, 1992). We found that in most cases, they receive remittances to their account for regular household expenses and other development activities while in some other cases, they receive a portion of the earning their husbands, who save the rest in their account. Decisions of these households are taken by mutual sharing between husband and wife by phone. In some cases, they have relative freedom to take independent decisions for everyday happenings whilst they receive instructions from husbands in other cases. Kusum (42) lives in a bhinno household with her son Suman (22) and two daughters Anwara (17) and Rahima (14), in the absence of her husband, who has been living in Malaysia for 15 years. One afternoon, while talking about her role in the household, she smilingly said to Laila and me, "My husband believes that I understand better than him. So, he allows me to make household decisions and go to bazaar, bank, or other public places." She further informed us that she goes to Sirajdikhan bazaar, 5 km away from Rashidpur, to buy rice, grocery items, spices, fish, meat, vegetables, etc., from wholesale shops. She goes to Sirajdikhan bazaar twice or thrice a month. She also buys a small number of retail commodities from the village bazaar, which sits every morning and afternoon.
However, Rashid (2013) argues that "husband is away" does not always mean that husband transfers his full power to his wife. Instead, he is constantly engaged in the process of household decisions through phone conversation. She further mentions that women do not necessarily enjoy their decision-making "power," "autonomy," and "freedom", while their husbands are abroad. But in the case of Rashidpur village, we found that, in previous times, although women would seek support from male relatives and neighbours to buy their commodities, the practice is gradually decreasing because the young generation wives consider themselves capable of managing the household keeping regular phone contact with their migrant husbands. They do not like commodities when other people buy them. They also do not know the actual price if somebody else brings their goods. Again, in many cases, migrants also do not want their wives to take support from neighbours. The migrants fear that if their wives frequently ask neighbours for help, people may spread a rumour of extramarital relations. Therefore, the wives consider it safer for them to move and purchase commodities by themselves.
However, we also found that although women manage everyday happenings by sharing with their migrant husbands without help from others, they maintain regular contact with their natal homes to seek help from parents or siblings when they need to. If necessary, they also ask for support from their husbands" father, brother, uncle or neighbours at different times. Through these male supports, they negotiate to create their position in patriarchal society. Roshna (45), the wife of a migrant to Singapore, said to Laila and me that she has three daughters and one son (14). While talking about her role in household management, she shared with me that: "I married our eldest daughter to a man of a nearby village by taking support from my brother, husband"s brother, and husband"s sister"s husband. They played the role of guardians who dealt with the groom"s party when they discussed marriage transactions, ornaments, and other arrangements. Now, the eldest son-in-law contributes as our son. Similarly, I married our second daughter in the same village, Rashidpur, with these male relatives' help. My husband provided financial support and necessary suggestions from abroad." The quote indicates that Roshna negotiates with trustworthy men to accomplish her purpose in the absence of her husband. She, further, said that when the proposal came for the marriage of the first daughter, she, along with her elder sister, secretly went to the groom"s village as strangers and saw him and his housing condition from distance. Male relatives also looked for the necessary information. She shared everything with her migrant husband by phone. The husband was not very willing initially as he did not see the groom by himself. Together with other relatives, she convinced him. But during the second daughter's marriage, he was convinced since the groom was from the same village, and he knew the family well. The husband also increased his reliability as the first son-in-law was a nice selection. Later, she requested her husband to bring the second son-in-law to Singapore. Their third daughter was married when her husband visited home for a few months. The eldest son-in-law runs a clothing store in Dhaka, while the other two live in Singapore near their father-in-law. The first son-in-law also supports her household work as her son in the village. Roshna represents many wives who actively take part in big household issues in the absence of their husbands.
When exploring women's property ownership, we found that many women try to save some amount from personal expenses. In some cases, young wives of ekloge households buy cow, goat, chicken, or duck with their savings to secretly rear in natal homes. They may also mortgage-in land in their father"s village and cultivate by the supervision of their father or and brother. Likewise, some women secretly open insurance accounts and once, considering the situation, share with their husbands. However, they stop saving in natal home when their children grow up. The savings in natal homes are temporary as they bring it back in cash when the households split or when the husbands need it.
The new wives' secret savings indicate that although they live in extended arrangements and conform to the household norms, they wait for their position. After the birth of children and with the growth of their age when they are confirmed about their position, they begin to concentrate on their husbands' households and decrease or stop saving in natal homes. These are their strategies to establish their strong feet in the households of their husbands. Gardner (1995) argues that in rural Bangladesh, women"s power depends on the ownership of resources. But, in my fieldwork experience, they have very little or no property ownership. They inherit a little property primarily from their fathers and secondly from their husbands. However, the findings of Rashidpur demonstrate that if they have a trustworthy conjugal relationship, they may achieve property from husbands as gifts. Rafida (34), for instance, informed that her husband, Nakib (45), was in Bahrain for six years and in the village for two years before he migrated to Malaysia four years ago. She lives in a bhinno household with her son (17) and daughter (14). Although her mother-inlaw (65) and brothers-in-law live in separate households in the same homestead, she has a good relationship with them. She asks for their suggestions before making important decisions. Initially, her husband used to remit to her younger brother-in-law, Anik (35), the most educated household member. She opened her bank account after the brother-in-law got married two years ago. As she maintains good relations with everybody, she has a very good reputation in the household. As a result, nobody protested when her migrant husband bought .30 acre of cultivable land for her as a reward for her contribution to the household and as the security of her old age. Rather the brothers-in-law negotiated when they settled the price with the landowner, who is their next-door cousin. Rafida"s case illustrates that a good relationship with the husband and other household members is an effective strategy that acts as social capital for women to acquire land (Bourdieu, 1986). It helps acquire property and creates an image in their favour to establish agency (White, 1992). Ortner (2006) argues that women's social relations and personality are the most important to establish their agency in the household and society. White (1992) supports this view that women who maintain good relation with their husbands and other members of the Uddin 7 household may be rewarded with expensive gifts, jewellery, land property, or home. The experiences of the women of bhinno households show how the absence of the migrant creates space for women to negotiate in the household decisions and move in the public sphere within the existing system. In many cases, the husbands also consider their wives" contribution as a big support to manage the household in their absence. Tenhunen (1999; in her studies on home-based wage work of women in Calcutta shows that culture not only restricts the agency of women but also creates space to enable it. She explains that when the husbands are engaged in white-collar jobs at the office, the wives secretly get involved in various wage works at home through intermediaries who sell the products in the market. These hidden works transform the women's cultural boundaries as they create social networks with their employer and the wider public sphere (Dannecker, 2002;Maharjan et al., 2012). Many husbands do not initially like the idea that their wives are earning, but slowly accept as the wives work at home without threatening their position in the household and society. Therefore, she argues that women"s agency is not obtained from complete denial of cultural norms but the creation of strategic spaces within existing gender codes (2006, p129-30). Sen (2018) also supports the creation of strategic spaces and shows in her study in two slum areas in India that although criticized for falling in love with Muslim men, Hindu women get the opportunity to express their agency by joining religious organizations in public domain and maintaining strong social network with others. Similarly, the wives of the migrant households of Rashidpur create their spaces through various activities without violating the norms and values of patriarchal society.

Women of the newly rich Bhinno households
The women of rich and middle-class households, usually, stay in homestead areas whilst the poor women are compelled to sell their labour and work in public places (Gardner, 1995;Tenhunen, 1999Tenhunen, , 2006. Many women of poor households are unable to use burqa and live religious lives. Their men allow them to move and work in public spheres. This practice remains effective even after the changes in their economic condition. Therefore, the study's findings show that the migrants who were poor before migration and now are prosperous because of remittance allow their women more to move in public places than the migrant-men from traditionally rich households. The findings also demonstrate that the women who were poor before migration and now are wealthy can reshape and redefine their social roles in their favour. These women have better negotiation and decision-making power in the household and in public sphere than the women from traditionally wealthy households. Contrarily, Rashid (2009) shows that the leftbehind wives of the migrant households remain stressed as they face the loaners and lenders every month in the absence of the migrants. However, in the case of Rashidpur village, although the poor women of many migrants, who went abroad taking loan or borrowing from others, remain stressed until repaying the loaned or borrowed money, enjoy dealing with cash and negotiation in household decisions when they do not owe to anybody. Their experience of dealing with loaners and lenders, despite stressful for a certain period, creates space to negotiate with outsider men and makes them proactive to some extent. On the other hand, the women of traditionally rich ekloge households have strong male support; but they reside under stricter norms and values, which keep them submissive (Uddin, 2018). In this respect, several women of bhinno households informed us that the women who have access to negotiation in household decisions and permission to move in the public places undermine the women of the traditionally rich ekloge households as they cannot move outside and depend upon their men even for their personal issues. They also informed us that the villagers respect the women of ekloge households because of their men's power and influence in the village. Similarly, they enjoy more public prestige because of their men; but as individuals, they cannot lead independent lives (Gardner, 1995). One afternoon in December 2017, Nazma (45), the wife of a newly moneyed migrant to Kuwait, said to Laila and me: "The women of rich and traditional households are unable to move alone. They cannot do anything without men. So, many women who can move in public places and make household decisions do not admire them." In another discussion, Bilkis (42), the wife of another migrant, said to us: "The men of traditionally rich households lead the village arbitration. But their women are restricted within the four walls of the households." Similarly, we found that many women of bhinno and newly prosperous households consider the covered life of the women of traditional households as ignorant and backward. So, they do not follow the women of rich and extended households as individual entities. On the contrary, they feel proud of their positions as they can support and contribute to household decisions or take many decisions alone. Since they are efficient in managing households, their husbands also rely on them. One evening, while hanging out in front of a tea shop, a returned migrant, Rashed (42), said to me, "Husband can increase their earnings staying stress-free abroad if the wife can manage the household alone." In another case, Rozina (35), the wife of a migrant to Qatar for 12 years, informed Laila that she had lived in a small corrugated tin house before her husband migrated abroad. They built a building two years ago, at the cost of approximately 30 lac BDT (35,000 US dollars). In the beginning, she was afraid and confused about how to accomplish such a big task. However, she engaged masons, carpenters, and labourers taking support from her brother. She regularly updated and consulted her migrant husband about the work through video calls and took suggestions. She gained confidence when the building was successfully completed. Now, her husband relies on her more with confidence. This woman is a representative case of how women of the newly rich household negotiate with men to manage the households, increasing their capacity to express agency.

Sustainability of agency and autonomy of women
We have found in Rashidpur village that bhinno households are managed by women by mutual sharing with migrant husbands and sometimes by taking help from close and reliable relatives. A few studies show that despite the material reward from their husbands' migration, women remain stressful as they play multiple roles in the household and in the public spheres (Seigmann, 2010). However, in the case of Rashidpur village, women are concerned about the food, cloth, and education of their children and the overall financial wellbeing of the household. Many men and women informed us that in most cases, women go back to their household duties when their husbands permanently return. Nonetheless, when their husbands are busy or when the husbands go to urban areas for a few days, they can use the experiences of going to bazaar, bank, or other public places. When the husbands are busy with regular earning activity like running grocery stores, clothes shops, tea stalls, etc., in the locality or work in urban areas, women go to public places to perform outside works like before while their husbands were abroad. In some cases, where women are employed to earn a living such as teaching or small businesses, they do not fully go back to household chores. Instead, they perform both in the household and in public places. Whether the women go back or not depends on the occupation of their husbands after migration.
Saleha (35), the wife of a returned migrant from Abu Dhabi, informed that she works as a representative of an insurance company where she has to move from door to door and convince migrants" wives to open insurance accounts. She has a very good reputation among the villagers. She is continuing her work as her husband supports her. She also informed that she was not always dependent on remittance to manage the household and her children's educational expenses when her husband was abroad. While talking about Saleha, a grocer (man) in the village bazaar said to me: "Housewives like Saleha are the assets for any household. They can support their family if the husbands are unsuccessful abroad or unemployed after returning." This comment indicates that many husbands expect support from wives when they are abroad or when return home. Similarly, in many other cases, it is not a matter of whether women go back to household duties or not; their experiences of managing the household in men's absence bring significant changes in their personality. As a result, they can actively participate in household decisions. In this respect, Rahela (38), the wife of a returned migrant from Saudi Arabia, said to Laila and me: "My husband lived abroad for a long time. Now, he runs a cloth shop in Sirajdikhan bazaar. As he spends the whole day in the shop, he does not have free time to buy retail commodities for daily needs. So, he likes to handover the responsibilities of household management to me." In another case, Mohibur (50) permanently returned from Oman, keeping his eldest son there. Now, he takes care of cultivation and fishing and stays almost the whole day at home. However, he makes household decisions sharing with his wife, Kabita (42). The wife said to me: "I managed the household for 10 years when my children were little, and my husband was abroad. I know what is necessary for the household and how to spend money effectively." She again said that her contribution is also not less to the improvement of the household. Her husband sent money, but she managed the household by applying her wisdom. If she could not manage properly, their financial condition would not be improved.
The experience that women gather in the absence of their husbands has a practical impact in the long run as it sustains after the migration period in many cases. Data show that women usually perform in the household and the public spheres to maintain constant contact with their husbands without support from relatives or neighbours. However, they take support male relatives and neighbours for big decisions such as marriage of daughters, buying land or building houses, etc. Thus, their husbands' absence creates a situation where they get rooms to negotiate and manoeuvre, which eventually help establish their agency and autonomy within gender codes (Jacobsen, 2011). Following Jejeebhoy and Sathar (2001), we consider that women"s decision-making power, mobility in public places, engagement with economic activity, property ownership, collection of information, and knowledge acquisition are the indicators of autonomy. However, they are not autonomous agents in the Western sense. Their positions are linked to various intersectional identities such as lineage, gender, class, and status in society when they deal with others (Cohen 1994;Brah 1996). Yet, they are conscious agents who can distinguish themselves from everyone else by expressing their Uddin 9 opinions and applying different strategies (Sökefeld, 1999). Therefore, migration has a lasting effect in shaping and reshaping the agency and autonomy of the women in Rashidpur.

Conclusion
The study has explored the diversity of the agency and autonomy of the left-behind women of overseas migrantmen. It has highlighted the continuity and changes in the discourse and practices of dominant gender roles in Rashidpur village, Bangladesh. It has revealed that women"s experience of managing households, transborder communication, and spatial mobility in the locality have created a situation where many women get rooms to manoeuvre within cultural codes. The study has shown that many new wives of ekloge households secretly save a portion of their personal expenses in their natal home whilst the wives of bhinno households maintain good relations with their husbands to create a position in the household. Many young wives of ekloge households do not take part in household decisions and rarely go to the public spheres, while the wives of bhinno households negotiate in household decisions and move in public places taking permission from their migrant husbands.
The study also has unfolded that among all the compositions, women of newly rich bhinno households, who were poor before migration and now are prosperous, are the ablest to negotiate, reshape and rearrange their boundaries and express their agency and autonomy within cultural codes. These women feel proud of their position as they can negotiate and move in the public spheres. The findings regarding the sustainability of the practices show whether the women will go back to their household duties or continue to perform both in private and public places depends on their husbands' occupation after return and their conjugal relationships. If the returned husbands are regularly busy with their occupation in the locality or urban areas, women need to perform in the public spheres like before. Similarly, if the women themselves are employed in earning activities, they do not fully go back to household chores; instead, they perform both in the household and the public places. The women who go back to household duties also can negotiate household decisions better than before migration. The study has followed the post-structural and intersectional feminist ideas and suggested that the agency and autonomy of the left-behind women and the system of their social relations are variously affected by the migration of men at individual levels. However, although they are conscious about themselves, they are not autonomous in the Western sense; rather, when they deal with others, their self and agency are connected to different intersectional identities. They negotiate, resist, or reinterpret their position to transform patriarchy and create spaces within the patriarchal system. Therefore, the managing experiences of women have huge social and cultural implications in the long run. In short, men's migration is instrumental in transforming gender relations in which women continuously strategize their positions as conscious agents.

CONFLICT OF INTERESTS
The author has not declared any conflict of interest.

ACKNOWLEDGMENT
The author would like to thank the ASA Conference 2018 participants who provided valuable comments and suggestions that guided this research. The author would also like to express deep gratitude to Carlo A. Cubero and Ellen Bal for their valuable comments and scholarly insights in finalizing the research paper, and the author"s research assistant, Runa Laila, for her tireless efforts and critical insights in the field.

NOTES
It is the lowest administrative unit of local government. Each union is made up of nine wards. Each ward consists of one or more villages. A Union Parishad consists of a chairman, nine members for nine wards, and three reserved women members who are responsible for supervising three wards by each. The chairman and the members are elected by direct election. Union Parishad is responsible for overseeing law and order and the government's development activities at the local level. It is a local government unit in the middle of Union Parishad and district.
It is an ideology that secludes Muslim women from the public sphere. It is a set of inner feelings and identity and a relationship between men and women. It emphasizes both the physical separation of women and also their need for male shelter (Papanek 1973). In Bangladesh, it is mainly maintained by burqa or veil.
According to Muslim inheritance law, the land property is transferred through the male line. Daughters inherit half of their brothers. Usually, they claim this property after the death of both parents and the father. In practice, in many cases, they do not claim this property as they have to take shelter of brothers if the breakup of their marriages occurs (Khan et al. 2016).