Gender Norms and Poverty Dynamics in 32 Villages of South Asia

The poverty dynamics of a community, and the social arrangements and opportunities that shape these dynamics, constitute important dimensions of well-being. This paper explores local understandings of and experiences with moving out of poverty and with remaining poor by employing the concept of gender norms, or the various social rules that differentiate women’s and men’s roles and conducts in society. The data demonstrate regularities in the influence of restrictive gender norms on understandings of poverty transitions, as well as how these norms are negotiated and bend to accommodate more gender-equitable practices on the ground. Our approach draws on feminist conceptions of gender norms that highlight their fluid and contextual properties, comparative case study methods, and a dataset of 32 village cases from five countries of South Asia. Villagers mainly associate movements out of poverty and chronic poverty with men and their capabilities to expand their earnings and assets despite limited work opportunities. Yet, our evidence from women’s life stories reveals examples from diverse contexts of women who exercise major roles in agriculture and actively work to improve the well-being of their families. However, these experiences rarely alter normative beliefs and practices that entitle men to control women and family resources.


Introduction
The poverty dynamics of a community, and the social arrangements and opportunities that shape these dynamics, constitute important dimensions of well-being and the freedom of people to lead the lives they value (Sen 1999). Various literatures concerned with well-being and development signal the contingent nature of how individuals and communities perceive well-being and freedom in their lives, with social norms and expectations, such as those associated with socio-economic position, gender or other markers of social identity, exercising important influences on perceptions (Ibid., Batz andTay 2018, Graham andChattopadhyay 2012). For example, the empirical literature finds that experiences with economic gains often foster a greater increase in perceived well-being among poorer populations compared to higher income groups (Diener and Diener 1995;Diener and Biswas-Diener 2002); however, as we highlight below, many studies have challenged assumptions about these linkages for women.
In this paper, we examine qualitatively the significant ways that gender colors local people's assessments of and experiences with moving out of poverty and with remaining poor. The sample spans 32 farming villages from five countries of South Asia. Our analysis of local poverty trajectories is framed by the concept of gender norms, or the "differential rules of conduct for women and men" (Pearse and Connell 2016: 35), to highlight the strongly gendered processes by which people perceive and experience transitions out of poverty and persistence of poverty. The analysis draws on feminist conceptual approaches that forefront the contextual, contested and fluid qualities of norms (e.g. Jackson 1998). The restrictive norms of this region present analytic opportunities to showcase the seemingly diverse yet "monotonous" beliefs and practices (Kabeer 2016) that slow the evolution of more gender-equitable poverty transitions.
The difficult nexus of rurality and the persistence of poverty and gender inequality remains a challenge in South Asia. A recent global panel study concludes that a child born into a poor family from the two regions with the highest concentrations of poverty, South Asia and sub-Saharan Africa, has a significantly lower prospect than a child born elsewhere of attaining a better life than their parents (Narayan et al. 2018). Data on ruralurban gender disparities in the region are scarce, but one study finds child marriage to be much higher in the countryside than cities of Bangladesh, and such practices raise women's risks of violence and exploitation (UNFPA and UNICEF 2017).
The influence of gender norms on the pathways and barriers to escaping poverty have mainly been addressed by case study literatures, which limits what we can learn from the larger comparative research of poverty dynamics (Razavi 1999;Ruspini 2001;Hulme and Moore 2010). In the following section, we introduce theoretical literature on gender norms and the concept's relevance for longitudinal studies from South Asia that address gender. Next, we discuss the methodology used to produce and analyze our dataset, which features women's and men's interpretations of and experiences with poverty transitions from diverse villages of South Asia. In the empirical section, we demonstrate significant regularities in normative influences on conceptions of poverty trajectories, including how they constrain and obscure recognition of women's economic initiatives in most but not all village contexts. We then conclude with reflections on the ways that norms discourage and accommodate-and, more rarely, become altered by-the initiatives of local women and men to move out of poverty.

Literature Review
Gender norms comprise socially constructed beliefs and practices that, across diverse cultures, associate men with authority and control over resources and women with submissive and reproductive roles. These norms include dictates that, for instance, constrain women's voice, freedom of physical mobility, and role in the economy, and represent an important social mechanism by which gender inequality persists (e.g. Seguino 2007). Relative to other regions of the world, the social conventions of the South Asia region interact in ways that often provide little latitude for negotiation. Patrilocal marriage and dowry traditions uproot girls and young women from their families and villages, marginalize them from family resources, and when combined with seclusion norms that constrain women's mobility, deepen their social isolation and economic dependence on husbands. The South Asia region also harbors the highest regional rate of intimate partner violence in the world (UNICEF 2018). Nevertheless, normative codes have fluid properties. For instance, practices associated with women's seclusion to their homesteads are often relaxed for poor women who must work (Drucza and Peveri 2018).
The available large longitudinal poverty studies that address questions of gender in developing country contexts consistently reveal women to be disadvantaged compared to men, albeit with some gender gaps declining such as in education (e.g. Narayan et al. 2018;Diwakar and Shepherd 2018;Van den Broeck and Kilic 2018;Baulch and Davis 2008;Quisumbing 2011a, b). An insightful mixed methods study, dating from 1996 to 2007 in Bangladesh, uncovered significant poverty reduction among communities participating in agricultural technology programs that targeted poor women, but mixed success with increasing women's asset control and reducing vulnerability due to events that deplete assets, notably dowry and wedding expenses and illness (Quisumbing 2011a, b;Baulch and Davis 2008). A recent mixed-methods study of poverty dynamics in eight countries of the Global South finds that agriculture remains an important pathway out of poverty; however, the authors highlight numerous adverse gender norms that constrain women's productive role such as, in the case of Nepal, restrictions on women's physical mobility, gender wage inequality, and women's "doubleburden of income and care work" (Diwakar and Shepherd 2018: 21).
Feminist analyses illuminate how gender norms operate and endure over time through mechanisms that are "both internalized (produces gendered selves) and also externally present and impinging through status expectations held by others and through institutional forces" (Sanyal et al. 2015: 18). Public surveillance and sanctioning practices play important roles in maintaining norms (Cialdini and Trost 1998). In many contexts of rural South Asia, and elsewhere, a woman who abides by codes of modesty and domesticity may be admired, while a woman who strives to be entrepreneurial with farming and move her family out of poverty potentially invites scorn, not only upon herself, but the entire family. Public actions to improve women's status, such as with inheritance and dowry laws, have been impeded by the durability of patrilineal kinship systems and the great value in these societies of stable and traditional family structures (das Gupta et al. 2004).
There is a longitudinal community case study literature from the fields of anthropology and sociology that brings to light the roles of values, norms, family hierarchy and other local level institutions in mobility processes to explain the causes of unexpected and sometimes contradictory effects of agricultural innovation and rural development on the ground (e.g. Rao 2008;Semedi 2012;Tickamyer and Kusujiarti 2012). A study of scheduled castes in two villages of Karnataka, India that began in the mid-1950s observed persistent gender gaps in both villages, even as one of the two agricultural villages prospered greatly (Epstein 2007). Women faced men's growing drinking and gambling, worsening domestic violence, and women councilors "still tended to act as the mouthpiece for their male sponsors rather than representing the demands of village women" (ibid.: 210). Another longitudinal study of two villages in Jharkhand, India, brings to life the diverse strategies employed by Santal women to defend their eroding land rights and gain recognition (Rao 2008). The women "had to tread carefully" and some risked their lives as they endeavored to garner allies among kin, community leaders, and courts to assert claims on property (ibid.: 36). Drinking men also appear in the Jharkhand work, and both studies highlight poor men's struggles with increasing marginalization as central to understanding women's heightened vulnerabilities with the penetration of markets, a decentralizing state, and other forces in the countryside (see also Farnworth et al. forthcoming). Gendered expectations for men to protect their families and provide adequately also exert pressure on subaltern men (Connell and Messerschmidt 2005;Jackson 1999Jackson , 2000. From the more nuanced analyses of local development and social change, moreover, it becomes evident that normative expectations vary on the ground and are not only constraints. Some gender norms provide for cooperative gender and family relations, and many dictates are continuously negotiated and resisted as women and men lead their daily lives and pursue their interests (e.g., Jackson 1998;Muñoz Boudet et al. 2013). A poor woman who generates assets may or may not be able to control these as she strives to move out of poverty. Gender norms mainly operate as constraining forces on women's agency and exercise significant influences on both women's and men's poverty transitions.

Methods and Materials
Our research draws from 32 community cases conducted in Afghanistan, Bangladesh, India, Nepal and Pakistan (Table 1) as part of the GENNOVATE research initiative (Enabling Gender Equality in Agricultural and Environmental Innovation). GENNOVATE's conceptual framework calls attention to how gender norms shape and are shaped by women's and men's agricultural innovation, and how these interactions potentially contribute to reducing poverty in a village and empowering poor women and men to improve their wellbeing. 1

Sampling Framework and Contexts
The study communities were selected based on maximum diversity sampling principles that specified variance along two dimensions: i) economic dynamism and ii) gender gaps. Economic dynamism was estimated using indicators such as infrastructure development, the integration of local livelihood strategies with markets, labor market opportunities, and resources available for innovations in agriculture. Gender gaps were estimated with reference to indicators such as women's leadership, physical mobility, education levels, access to and control over productive assets, and ability to market and benefit from sales of agricultural produce.
The communities differ in numerous ways. The villages are populated by diverse caste or tribal groups, and several report both. 2 With minor exception, the Muslim religion prevails in cases from Afghanistan, Bangladesh and Pakistan; Hinduism in India; and Buddhism or Hinduism in Nepal. 3 Weather shocks and political conflict disrupted lives in some villages from all five countries, and insecurity persisted across all four Afghan cases. In one remote village, from Pakistan, community members testify to near feudal-like conditions, with most village workers, including children, shouldering labor obligations to landlords due to protracted indebtedness. Elsewhere, the literature and our own data point to less rigid socio-political conditions, as rural power structures have become more diverse with the reach, albeit uneven, of markets, roads, public services, political parties, international and nongovernmental actors, and an array of more and less formal local networks, including women's own self-help groups (e.g. Lewis and Hossain 2017;Epstein 2007;Narayan et al. 2009). Irrigation is present in all but four of the 32 case studies, and electricity is largely available except in most villages visited in Afghanistan and Bangladesh. Most boys and girls are indicated by key informants to be attending primary school in the study villages, except in three Pakistan cases. Most children also attend some secondary school, except for in the Pakistan cases and girls in Nepal. Data Collection Methods GENNOVATE's data collection tools (Petesch et al. 2018a) are inspired by participatory methods that enable women and men to reflect on and interpret their own lives and experiences. In each research village, data collection included six sex-specific focus groups with (i) poor women and men, (ii) middle-class women and men, and (iii) young women and men. In addition, nine semi-structured interviews were conducted guided by three instruments: a) a community profile to gather background demographic, social, economic, agricultural, and political information about community (requires key informants of both genders); (b) innovation pathway interviews with local people who are known for trying new things in agriculture (two men, two women); (c) lifestory interviews (two men, two women). Table 1 highlights the primary data used for this paper, which include 64 focus group discussions with poor men and poor women and 39 semi-structured life story interviews with women. Due to space constraints, decisions were made to explore and compare two types of poverty transitions among the set of women's life story interviews: those who move out of poverty (12 women) and those who remain poor (27 women). The categorization of the mover and chronic poor sets was done by combining the reported assets and livelihood activities with responses to a question about how the woman's household is getting by with the current level of income. Movers, for example, consistently testify to getting by with ease, while chronic poor responses convey struggles, such as with periods of hunger; loans taken from relatives, shopkeepers and others. The women's ages range from 28 to 54.

Analysis
Our approach to qualitative comparative analysis involves working iteratively with two analytic procedures (Miles et al. 2014). The first employs variable-oriented measures that draw on systematic content analysis methods and GENNOVATE's coded dataset (in QSR NVivo, a social science software) to identify recurring themes across cases associated with the mobility dynamics captured in the focus group and life story testimonies. The coded dataset enables analysis of, for instance, the different types of livelihood activities, family relations or networks associated with the testimonies that depict upward or chronic poverty trajectories. The second approach to analysis is the contextual case-oriented work that focuses on the gender norms associated with mobility experiences in a specific village or person's life, and how people negotiate these expectations.
We recognize that findings are inherently colored by status differences and quality of relations between the researcher and research subject, and the significant interactions between data collection, interpretation processes, and junctures of time and place (Feldman and Welsch 1995;Portelli 2016). In diverse ways, what study participants elect to share and how researchers grasp and interpret what is shared are colored by questions of position and distance. Petesch et al. (2018a) provide additional discussion of GENNOVATE sampling, data collection and analysis protocols, including issues of recall, courtesy bias, informed consent procedures, translation, coding, data triangulation, research ethics, and other common field research concerns.

Findings
Focus groups with poor villagers widely perceive men to exercise the principal role in the poverty trajectories of their families and villages. Men's initiatives dominate the discourse even though we also probe directly on women's experiences with getting ahead. Men's roles grow even larger, moreover, in testimonies about why families stay poor. The analysis then shifts to a set of women's life stories to showcase the fluid properties of norms, and the women's substantial roles in the mobility processes of their families and villages.
Across the 32 study communities focus groups observe a median of roughly 60% of their village households to have been poor a decade ago (or 2005/6). 4 Over the ensuing decade, women's groups indicate poverty to have fallen to 38% while men perceive nearly half of their village population as still poor. When comparing trends by country ( Fig. 1), women's groups from India and Nepal observe the strongest poverty reduction (40%), and men of Afghanistan the least (9%). In most case studies where women indicate greater poverty reduction than men, this was because the women's focus groups perceive households with cropland tenure or workers with stable jobs to qualify as no longer poor, while men's focus groups often identify all manner of smallholders, shopkeepers, low level civil servants, moto-taxi drivers, and so forth, to be poor because the earnings from these jobs are unable to provide adequately for their family's daily needs. As women rarely own land or hold stable jobs in these villages, their markers for distinguishing poor households from others strongly center on men's economic capacities.
Our evidence showcases diverse gender norms that constrain women's productive roles and privilege men. In most villages, women's presence in the public sphere is strongly discouraged, and they frequent the market rarely and only if accompanied by a male relative. Women often testify to needing their husband's permission for them to generate income. In most villages from Afghanistan, Bangladesh, India and Pakistan, women with produce or handicrafts to sell conduct their marketing from homesteads or with the help of family members. Nevertheless, we highlight below some of the examples in our data that signal flexibility with rural seclusion practices and women's productive roles.
The three Nepal case studies stand apart from the others on some key dimensions that shape gender roles and resource access. At 83%, women's labor force participation rate in Nepal is very high by global standards, and even more so compared to rates of between 20 and 33% among the other four study countries. 5 Fig. 2 presents estimates provided by local key informants on the share of women's day laboring on farms across the case studies in 2005 and 2015, revealing a dramatic increase in Nepal following the end of the country's civil war in late 2006. In addition, women from Nepal report much greater freedom to be mobile in their villages relative to the other cases. 4 The poverty levels are based on an exercise that engages focus group members in building a ladder that depicts traits of the different socio-economic groups of their village at each ladder step. Then the group establishes the ladder step at which households are no longer considered poor in their village. Next they sort 20 seeds that are representative of all the households in the village on the different ladder steps; and the seeds are sorted again to represent the distribution 10 years ago. For an example of two ladders from a similar study conducted in rural Andhra Pradesh, India, see Narayan et al. (2009: 244). 5 World Bank DataBank based on ILO model; accessed January 27, 2019.
Figure 2 also reveals that many poor women engage in daily wage farm work in the India sample. Especially in the cases from India, but also elsewhere, study participants often convey a confusing mix of norms that alternately discourage and accommodate women's income earning. These types of testimonies speak to the fluid properties of gender norms, and how they loosen or tighten, and occasionally disappear, even deeply rooted expectations like "only men are farmers." Farnworth et al. (2018) present a case from the Nepal sample to demonstrate how women's roles are shifting from working on farms to actively managing them and employing the latest agricultural technologies and practices, often supported by and in consultation with husbands in distant jobs. Similar dynamics of women stepping into more managerial agricultural roles can be seen in the life stories from the women in our sample who moved out of poverty in the other countries as well. The experiences of these women, however, appear relatively infrequently in testimonies from focus groups about the poverty dynamics of their village. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% India (12 cases) Nepal (3) Bangladesh (6) Pakistan (7) Afghanistan (4) (12) Afghanistan (4) Bangladesh (6) Pakistan (7) Median % of village women Now Ten yrs ago

Gender Norms and Poverty Escapes
In each study community, a group of poor women and a separate group of poor men discussed how the men of their village have contributed to moving their households out of poverty, and then they discussed the women's contributions. 6 Revealing of the strong gender differences in this discourse, the word count from the men's responses to the question about how men escape poverty (4615) proved to be nearly double what women had to say about how women accomplish this (2428). 7 "Women usually cannot bring a big change, but they can assist their men in going up," explains a member of the poor men's focus group of Ismashal village 8 in Pakistan's Khyber Pakhtunkhwa (KPK) province. Figure 3 below presents the leading topics that emerged in responses to the questions on how men and women move out of poverty. The frequencies highlight that focus groups largely stress similar dimensions for both genders. To get ahead, men and women alike employ varying combinations of agricultural activities, nonfarm jobs, asset-building measures and persistent hard work, the latter of which was coded to "emotions, behaviors, attitudes." Beyond these leading dimensions, it is noteworthy that men almost never mention their spouse (marital roles) in reference to men's upward mobility. Yet, when the topic turns to women's efforts to move up, women's focus groups, and even more so men's, testify often to how couples cooperate, for instance, by women supporting their husband's farming by laboring in the field with him or provisioning cash from their livestock activities.
And while the mix of factors appear quite similar for both genders, the coding frequencies conceal perceptions of considerable gender differences. Focus groups overwhelmingly associate men with significant opportunities, such as purchasing land, managing their own commercial farm, opening shops, working in government and private sector jobs, engaging in labor migration, taking large loans, and so forth. Women, by contrast, are mainly depicted as engaged in smaller initiatives to get ahead, such as supporting their husband's farming in villages of Pakistan, tending to livestock and vegetable crops from homesteads in cases from Afghanistan and Bangladesh, taking up farm jobs in India and Nepal, and tailoring across diverse communities. Additionally, women's careful stewardship of savings often emerges as an important strategy: "If a man gives 10 rupees to a woman, she will save two even from this small amount" (poor women's focus group, Murmura village, Bihar, India).
In Nepal, and a few cases elsewhere, nevertheless, we find reports of women engaging in commercial farming much like men. In Pakadi village of Nepal, women say, "We grow both seasonal and off-season vegetables and hybrid varieties of crops to make more profit." Similarly, in Deva village of Uttar Pradesh, India, the women's 6 As discussed in note 4, these questions follow detailed discussions on the socio-economic conditions and trends of the households in their village. 7 Please note that word counts differ from the frequency of mentions conveyed in Figure 1. A "mention" refers to a text passage that has been coded to single topic, and a passage can range from a brief phrase to a set of paragraphs. Also, some of the disparity in word count can be attributed to the women's focus groups, and less so the men's, having already discussed spouses and other family members in the initial question about men. 8 All names of villages and individuals are pseudonyms. focus group observes little difference in how men and women escaped poverty in their village: People on step 1 (10 years ago) have worked very hard to rise above poverty and earn enough to be able to live decent lives. Both husband and wife work hard in farm labor and other casual labor to be able to earn as much as they can. Some of them started sharecropping and others took land on lease. They have saved whatever they could and bought small plots for additional means of livelihood. Hence, they have slowly risen to step 2.
Pakadi and Deva also happen to be contexts where women report extremely high rates of poverty reduction (40 and 67%, respectively), and their testimonies indicate that norms have relaxed in ways that support women's initiatives to engage in and benefit from commercial farming opportunities and contribute to their villages' prosperity. Another community from Bangladesh, Borian, near Mymensingh city, presents a case where local women are returning from urban factory jobs, buying land and raising cattle and goats. However, focus groups elsewhere rarely talk of women engaged in such significant initiatives.
The heavy coding to emotions, attitudes and behaviors brings to the surface the strong agency and psychological dimensions associated with upward movements. To get ahead, both women and men repeatedly stress a need to be "hard working," "struggling hard," "take a lot of risks," "persevere" and "fight against all odds." Particularly among the Pakistan testimonies, but also elsewhere, women and men recognize the emotional labor that women shoulder for their families by keeping "her husband away from stress," supporting her "husband in all of his efforts," not making "unnecessary demands," and being "content with his income." By comparison, men's emotional support to their wives is rarely mentioned. Such narratives attest to the highly normative pressures on poor women to run the household adeptly on scarce resources and on poor men to provide despite the unreliable, arduous and low paying work available (e.g., Jackson 1999). These stresses around the patriarchal bargain move us  directly into the even more strongly gendered testimonies about the drivers of chronic poverty.

Gender Norms and Chronic Poverty
When a household persists in poverty, focus groups largely attribute this to men's challenges with income generation and willpower. In the face of a man who cannot find his way forward, a poor woman is widely perceived to have little room to maneuver. Figure 4 presents the main topics that surface in the focus group responses to the two follow-up questions about why a poor man, and then a poor woman, is unable to move their household out of poverty. First and foremost, men and women identify a lack of work opportunities for men to generate enough income to maintain the family as the leading reason why families stay poor. There are testimonies specific to agricultural hardships, such as the meager and unreliable returns from farm jobs, problems of landlessness and unproductive land, and weather shocks, but mainly focus groups stress that there is "never enough money" to cover expenses.
Many also consider poverty to be a consequence of men's own making because they are "not that smart," "lack ambition," "lazy," "don't work hard," "idle," and "enjoy [life] and stay home until the money runs out" (women's group, Rawatgaon in Rupendehi, Nepal), or "shy away from working hard and long hours" in daily wage labor jobs (men's group, Bete, Punjab, India). Men, as well as adult sons, also stay poor by squandering money on "bad habits" like drinking, gambling, drug use and other illicit activities (the vices coding). "Some men say they drink in despair"; and their wives fight with them because "they are making a bad situation worse…" relate members of the women's group of Cheeda, India. The women of Cheeda talk of poor men beating their wives as a driver of chronic poverty.
Poverty persists due to myriad family circumstances (the coding to housework, parenting and care roles) that include many children, many daughters and dowries, lack of help from (idle and lazy) adult sons, and care of those who are ill, disabled or elderly. These needs are compounded by a poor family's lack of money and other assets to access better work or to cover expenses, and their indebtedness to local shopkeepers and moneylenders. Shocks due to illness and dowry are frequent findings in the longitudinal literature about triggers of descents into poverty and chronic poverty (Quisumbing 2011a;Baulch and Davis 2008); however, our broad coding categories, such as lack of money or household care needs, Fig. 4 Leading topics associated with remaining poor (GENNOVATE coded dataset, 64 focus groups). Figure note. All but three of the references in the vices coding to why women stay poor are in reference to men's drinking, drug use, gambling or illicit activities. flatten findings from more specific codes for illness and weather shocks, and how adverse events often combine. These processes become more evident from the life stories below.
The coding patterns mirror those for upward mobility in that references to spouses emerge frequently as explanations for women's chronic poverty but rarely as causes for men. In Shanti of Uttar Pradesh, India, a woman expresses a common refrain in these testimonies, "Women suffer their husband's fate." This was especially seen to be the case if the husband was not actively working to maintain the family; and in many of these communities it continues to be acceptable to berate and physically punish a woman who pressures her husband to provide or who seeks to generate income of her ownas these moves challenge men's authority.
Focus groups sometimes consider women who themselves fail to build savings or who make unnecessary purchases as reasons why households cannot get ahead. Women also remain trapped because of other gendered expectations that make them "mostly busy with household chores" (Panali, in Kabul Province, Afghanistan); and "they cannot work outside the home for fear of losing their reputation and respect" (women's group, Master Goli, Rangpur, Bangladesh).
Contextual differences emerge strongly in these testimonies. In focus groups from KPK, Pakistan, women say they remain poor because they are "almost slaves," "depend on men", and cannot be involved in moving their households ahead because they do not have any jobs, land or money. In Afghanistan, men and women attest to scarce opportunities due to the continuing conflict, distrust, poor economy, rampant corruption and need for special connections, and the expectation of women's subordination to men, e.g.: "what can a woman do to make him go outside and work?" (men's group, Lehsat, Afghanistan). Even in contexts where large numbers of working men are away, and many poor women must farm independently, some still testify to different normative constraints on their capacity to generate income: "[W]e can see that families are doing better where women have started working; yet mindsets don't change" (women's group, Murmura, Bihar, India). Whether reflecting on climbing up from or being stuck on the bottom ladder steps, women and men stress obligations on men to be men and to take charge and provide, and how very much poor men and women invest every day to negotiate these expectations.

Some Move, Others cannot
For the remaining analysis, we turn to the women's life story dataset to explore the normative dimensions of mobility processes in more nuanced ways than is possible with focus groups. In these testimonies, the fluid properties of gender norms become apparent as diverse types of women explain and interpret the trajectory of their lives and wellbeing over the past decade. A key dimension that sets apart most movers is how they secured productive assets from husbands or brothers that enabled them to begin processes of accumulation and innovation with their agricultural livelihoods. Family relations also figure significantly in chronic poor trajectories but mainly to support the women in their gender-ascribed reproductive roles or to cope with shocks.
When comparing women in the mover with the chronic poor set, similarities can be seen on several dimensions. As shown in Table 2, both on average have little education, are mothers of four or five children with the youngest around ages 10 or 11, and most live in homes with secure tenure. 9 Moreover, more than half in both sets report experiences with public assistance, a family member's labor migration, and a major family illness or accident. Both mover and chronic poor women similarly report combining and moving in and out of varied livestock and field agriculture activities. In India and Nepal, most women's occupational histories included day laboring on farms, and many women from all countries testify to experiences with sharecropping and leasing land. Reports of domestic work, tailoring and daycare jobs are also present but less common, and women often combined these jobs with agricultural work.
Notwithstanding the many commonalities, the mover set differs in significant ways. These women are much more likely to be married, belong to a landowning family, and manage large livestock. Most movers, moreover, are employing modern agricultural technologies and practices: -In 2015, using zero tillage machines I started maize farming, for which I had a great yield and large profit …. (Mover, Matipur, Bangladesh) -In 2005 I started my cultivation work, like growing betel leaves. And owning cows and goats. At that time, my financial situation starts getting better. I used to buy cows on Eid-ul-Fitr and sell them and buy more cows and own them. (Mover, Dampur, Bangladesh) -I work on my own land as well as on sharecropping. Since I started using zero tiller there is not much field preparation to be done; so my work has become easier and In ten of the 12 mover's life stories, the women are engaged in commercial farming activities on land that they themselves acquired or significantly financed the purchase. The remaining two movers are generating income from dairy activities. One of these two is an Afghan woman in her fifties with an elderly husband. She used to work extensively in the fields but currently focuses on dairy activities, as she has adult sons who now take care of the croprelated activities. With loans from her brothers and a cousin, she leased the cropland and launched a tractor-service rental business for her sons to run. Movers are also somewhat more likely to reside in extended families. In the next section we illustrate further the importance of women's perseverance with their livelihoods and negotiation of family resources. By comparison, chronic poor women rarely mention significant benefits from their farming opportunities or employing machines or other modern technologies and practices. But underpinning these differences are social processes with important regularities. More than one-third of the chronic poor women are widows, separated from or abandoned by their husbands. Five others testify to serving as their family's main provider due to husbands who had problems with drinking or violence, or both. Another three women, all in the India sample, run their household due to husbands who are disabled. The chronic poor also testify to trajectories stymied by varying combinations of family health crises, displacement due to conflict or floods, daughters' dowry obligations, failed migration experiences, and crushing indebtedness. These hardships are present in the movers' narratives as well, but do not derail these women's momentum with their livelihoods. Our evidence strongly suggests that an important reason is that the movers had an influential family member who was willing to flout local customs and provide a woman with economic support to expand their agricultural activities.
The movers are also more likely to enjoy civic ties, such as participation in self-help and informal credit groups and training opportunities. Whether mover or chronic poor, however, women rarely mention social ties beyond the family unless prompted. Instead, it is family bonds that emerge most often. Towards the end of their interview, we ask the women to reflect on the most significant relationships that have helped them to pursue important goals in their lives. Half the movers named husbands in their responses, followed by brothers in another quarter (Fig. 5). By comparison, chronic poor women reach further afield to find support, if they identify allies at all (Fig. 6). It is more common for chronic poor women to remark that "there hasn't been anyone that important in my life" (chronic poor, Shanti, Uttar Pradesh, India) than a husband who "has always been there for me" (chronic poor, Prem, Bihar, India). Sons or daughters are identified in about equal measure to brothers by chronic poor women. By comparison, movers never mention children in these responses. For movers, husbands are most often "my pillar of strength" (mover, Prem, Bihar, India), while, for chronic poor women, husbands could be "the source of all my worries and sorrow" (chronic poor, Cheeda, Uttar Pradesh, India). Or, a brother could be "very cruel and never consider me his sister or even a human being" (chronic poor, Katam, Nangahar, Afghanistan) or the very reason why "I achieved my goals and became prosperous in life" (mover, Ismashal, KPK Province, Afghanistan). The importance of these various relations to the women's trajectories comes into more focus next.

Norms Relaxing and Tightening as Roles and Relations Evolve
Here we present and compare the life stories of a mover and chronic poor woman in each of two villages. The first two women reside in a large and dynamic village of western Bangladesh and the second two in a smaller and remote community from Bihar, India. Both villages are making progress on poverty reduction, but focus groups estimate roughly 40% still to be poor in the first case, and 50% in the second. While the life stories cover a broader set of concerns, the emphasis here is on the events and relations associated with individual women's capacities to strengthen their livelihoods and accumulate assets and that the women considered to be important for understanding the ups and downs of their lives and wellbeing over the past decade. As the gender hierarchy prescribes, both mover and chronic poor women often framed their trajectories in relation to important men in their lives.
Dampur village hosts a growing population of 4000 and resides in western Bangladesh in Rajashahi district. Improved transport, cellular service, irrigation and other infrastructure development help farmers to make good profits on their rice, vegetables, jute, betel leaves, wheat, maize and fish. Key informants also report various governmental and NGO programs that have supported local farmers to intensify crop production and diversify into aquaculture. The two women who shared their life stories report that they did not participate in any networks or extension opportunities, and the young women's focus group indicates that a woman would not be welcome at a community event where agricultural information is provided. The village's processes of agricultural innovation, in other words, have mainly reinforced the community's restrictive normative climate for women's economic agency.
Age 40 and with two children, Sufia is the mover from Dampur. When Sufia and her husband were expecting their second child, her in-laws pushed them out of their home due to unspecified conflict. However, Sufia portrays her husband as "a little lazy"; and one suspects that his parents might have considered it past time for him to step up and provide, and for them to feed fewer mouths. The couple coped by building a house on government land. Especially interesting, however, is that Sufia used this transition into a nuclear family as an opportunity to gain access to a plot (11 kathas) with the backing of her brother and father. Further revealing of the conflict surrounding this couple and the strong influence of gender norms, Sufia's husband refused to allow her to be named on the title for the plot purchased by her brother: My husband said if the land is not written in my brother's name then he will divorce me…. I agreed because I didn't want to cause any trouble… But I took the decision of cultivating the field, and my husband agreed to it. And from the produce I could buy other land with that money.
Ejected from his family and landless, the husband likely perceived his authority in the family and more widely directly challenged by his in-laws' gift of land to his wife. Questions about family violence were not posed during the interview, however, in their testimonies, the poor women's focus group observed a sharp spike in Dampur's women being beaten in family conflicts over the past decade, with the rising costs of raising children indicated to be a major driver. Given the village's strict norms that constrained women's income generation and the strife implied by Sufia's testimony, it must have been deeply challenging for this young pregnant woman to press forward with her own agricultural initiatives. Nevertheless, Sufia poured herself into bringing the plot to life. Along with raising goats and cows, and the housework and care needs, Sufia cultivated betel leaves in 2005 and added paddy the next year. By 2011, her farming proceeds enabled the purchase of more land and she diversified into eggplant, chilies, and bitter gourd. Sufia says she learned to cultivate these crops from her grandfather when young. In 2014, Sufia qualified for a major loan to set up her son in a grocery shop, "From which my family started earning more." 10 She further adds that the son is responsible for repaying her loan. Sufia reveals later in her testimony that her husband suffered a stroke in 2006 and required years of medical treatments until he passed away in 2013. When Sufia reflects on her life, she considers the most important relationship to be with her brother, "because of him I can now stand on my two feet." Next we turn to Renuka, who also lives in Dampur and remains poor. Renuka's husband divorced her shortly after marriage at age 15, and she returned to her family. She never remarried, is now 55 years old, and caring for her mother in a small home on her middle brother's land. She says most every family and community member shuns her. Renuka had inherited a small plot from her father but she turned that over to her brother, who shares some of the produce. She mainly gets by with seasonal farm jobs and income from livestock, which she recently expanded to seven goats and ducks and 15 chickens. A sister lent and then gave Renuka a goat in 2007 after she faced periods of hunger in 2005 and 2006; and as her livestock grew, Renuka says this enabled her to start generating savings. At this juncture, Renuka is resigned to her isolation and poverty, and concludes that her life would have been easier if she had land: "I could have earned a lot of money from farming my land." Like many women who inherit plots, she needed her family's support and protection rather than alienating them by hanging on to the property. The various gendered expectations that disempower women with no husbands reinforced Renuka's chronic poor trajectory.
To examine the next pair of life stories, we move to Murmura of Bihar, India. With a population of 500, Murmura is predominantly Kurmi caste, and most village workers take day jobs on farms or sharecrop. There is no electricity, nor the extension and NGO services of Dampur. Both Murmura's women and men produce dairy and farm in the surrounding fields of wheat, paddy, maize, pulses (dalhan), and dhencha, (a manure crop). Yet, few women move independently in the village. "Girls live under a lot of restrictions to save our 'pride' and the pride of the family," explains a participant of the young women's focus group. Murmura's focus group with poor women report a rise in domestic violence compared to a decade ago, but not at the rate of Danpur's.
Jana, age 43 and mother of three daughters, heads one of the many families of Murmura that remains trapped in poverty. When the youngest was three, Jana's husband left to find better work to support the growing family. He was never to be heard from again, leaving Jana the weight of raising and marrying off three daughters. Jana initially fed and educated the girls by taking loans and farm work; and over time she managed to combine farm jobs with sharecropping, "which allowed me to start saving." But Jana says one of her biggest challenges is that her yields are a tenth of what they could be because she cannot afford irrigation or fertilizer. She considers further debt to finance her farming too risky. In addition to small loans from the shopkeeper, Jana mentions borrowing over the past decade from the moneylender for her second daughter's wedding in 2010, to improve her dilapidated house when it fell apart in 2013, and to cover medical expenses when she became very ill in 2014. Her inlaws gave her the title to the house, and she mentions a brother helping with medical expenses. But to maintain the household, much less move out of poverty, she has only herself, and for a few more years, a 12-year-old daughter who works with her in the fields. Jana mentions friends in the village who are proud of how she has managed independently, but she worries greatly about how she could cope with another illness and aging in the years ahead.
Finally, we turn to Anita of Murmua. She is 54, a farmer and mother of eight, and her household only recently moved out of poverty. Anita opens her story by sharing that she married into a poor landless family and took her first job 25 years ago after losing a sick child: "… there wasn't enough money to take him to a better doctor. This jolted me … and I decided to go cut turai (ridge gourd)." The decision incurred great wrath: … my husband was very angry and did not eat for two or three days because he just hated the very idea of my working on another person's field. Then I talked patiently to him and made him understand my viewpoint. I told him that if he went to work in town he would earn more, and if I worked simultaneously here in my village, I would also contribute to family funds and we would be able to save some for emergencies. I reminded him how we had lost a child for lack of money. Finally, he relented.
The women's poor focus group of Murmura cautions that in their village "there is fear about what others will say if the wives work," and indicates it to be rare for a young married woman to work for pay. One imagines that 25 years ago Anita's actions caused a scandal, and her husband had reason to be concerned for his own and his family's reputation. In fact, Anita confides that "People definitely spoke behind my back and made fun of me because I worked in other men's fields." Once she had her family's support, however, Anita says she no longer worried about "these petty issues." Other women interviewed similarly testify to processes of great resistance to and then acceptance of their income earning, and over time, gaining greater respect and say in the family.
For much of the marriage, Anita's husband worked in a distant town and sent remittances. When he returned home to stay over ten years ago, however, the large family was still quite poor. Anita remarks that her "husband cooperated fully" with scaling up their farming activities, but her narrative clearly indicates that she led the way. She explains that in 2005, "I leased 5 or 6 kathha of land and started working on it since my husband was never comfortable with the idea of my working on other people's plots." In order to lease the land, "I sold all my jewelry, in consultation with my husband." In 2010, she was able to lease additional land and her savings continued to grow, so that by 2014, "We finally managed to purchase our own 8 kattha plot. It was a big achievement." Anita owns the land jointly with her husband; and even at age 54, and with five sons, she still works every day in the fields. They have a loan to repay from a daughter's wedding, major expenses that include a son who is studying in Chennai, and debts from her own and her husband's illnesses in 2014. Had normative conditions and economic opportunities in Murmura been more inviting for women like Anita, one imagines she may have gotten her family across the poverty line much sooner.
While the social conventions that shape access to assets and transfers of family wealth strongly privilege men, the testimonies make evident that women are actively engaged in sharecropping, leasing, and acquisition of farmland. This was the case whether a more dynamic village of Bangladesh or a more remote one of India. The women's experiences, nevertheless, reveal significant politics and strategic timing to accompany whether and how a husband, brother, or other kin will invest in them to accumulate livestock or to manage cropland and succeed economically. Women who can mobilize enough family support to launch their own commercial farming and then bring sons along appear to have significant advantages. Sometimes other opportunities can help mobilize the necessary acceptance, if not outright support, of family. A 37-year-old mover from Matipur, Bangladesh, credits her participation in the Union Federation women's farmers group with enabling her to turn her life around due to training she received on paddy and fish farming. Mainly, though, movers stress their own perseverance as well as the support and resources that they mobilized from husbands and brothers. Chronic poor women also strived to get ahead but lacked kin or other means to bolster their productive projects.

Discussion
Across the 32 village case studies, prevailing understandings of poverty transitions emphasize men's roles and devalue women, even as women, in addition to maintaining the household, also generate income and assets that lift their families out of poverty and help them cope with shocks. The persistence of such patriarchal framings attests to the powerful capacity of gender norms to relax or tighten to accommodate the changing circumstances of women's and men's lives without requiring a similar change to the deeper values, family structures or other institutions that regulate the social life of families and communities (Sewell Jr. 1999). The contingent nature of norms sheds light on important social processes that impede the transition to more equitable norms as households and villages become more prosperous.
Villagers overwhelmingly associate men with the means to amass assets, command family labor, travel to distant jobs, and bring their families out of poverty. Processes of chronic poverty are even more tightly attached to men, and their struggles with provisioning and fulfilling their gender-ascribed role. By comparison, when focus groups reflect on women's roles in the mobility processes of their villages, their contributions are acknowledged, but women's fates are widely seen to be tied to men's, which conforms to dictates that emphasize women's dependency. The testimonies speak to pressures on women to restrict their livelihood initiatives, avoid family conflict, and bolster men's self-esteem, which are common findings in the gender literature and signal some of the ways that women, too, uphold gender hierarchy in their daily life (e.g. Jackson 2000; Anderson and Jack 2016).
Additionally, women and men sometimes act in ways that restructure rather than preserve the rules (e.g. Muñoz Boudet et al. 2013). In Madpur village of Faridpur district, Bangladesh, poverty has fallen faster than in most study villages, and women there testify to a normative climate where it has become acceptable for them to be earning income, such as by combining livestock activities, vegetable and fruit production, and diverse daily wage jobs. While women lament how much longer they work every day than their husbands, they still want the changes underway in their lives to continue: "We think the matter of equality between men and women is good for all of us." Poor men in this village express similarly favorable views about the changes in women's lives and relate how closely couples cooperate to move their families ahead. In most villages, including those experiencing significant poverty reduction, there is much less evidence of norms that support women's roles in the economy and cooperative gender relations.
When we turn from focus groups to life stories, women's agency emerges with great force as they recall their family relations, livelihood initiatives, and other key economic and social experiences shaping the trajectory of their lives. The women often testify to negotiating pressures not to work for pay, not to make claims on family resources, not to challenge men's authority in their efforts to undertake or expand their livelihoods. Movers and chronic poor alike also attest to persevering through periods of profound insecurity due, for instance, to marrying as a child and leaving their natal home, illness and death of loved ones, family conflict and separation, crop and livestock losses, and deep indebtedness. The women who made their way out of poverty took great reputational and financial risks with securing the consent and material backing from husbands, brothers or other relations in order to scale up their own agricultural activities. Among the movers with adult children, many shared proudly of marrying off daughters while investing in cropland, shops and higher education for sons-processes which contribute to the intergenerational production of gender inequality. By comparison, chronic poor women in our sample report scarce or no support for their livelihood activities, even as most had become the primary or sole income earner for their family. Nor could they do much to bolster their children's fortunes. Across diverse types of villages and households, patriarchal norms prevail that entitle men to control over women and family resources and contribute to weakening the contribution of poverty reduction to gender equality.

Conclusion
Gender norms exercise significant influences on poverty transitions. Across the study contexts, the possibility of families to move out of poverty is deemed to depend on men's roles. This is the case even where many poor women toil in agriculture to build savings, educate children, provide secure housing, and acquire cropland, or whether they pawn their only assets to deal with crisis. The latitude for women to work for pay and build productive assets varies across the cases, as well as within the villages. Over time some women have negotiated substantial roles in small-scale commercial agricultural enterprises and moved their family out of poverty. In relatively few villages, however, are such practices widely recognized or encouraged.