Rethinking Rural–Urban Migration and Women’s Empowerment in the Era of the SDGs: Lessons from Ghana

: Women who migrate within national borders in Africa have been largely ignored in contemporary conversations about migration. This is partly due to the fact that internal migration, and in particular, rural–urban migration, has been viewed in a negative light in development theory and praxis. This leads to the perception that women who migrate within national borders are worse-off than they would have been otherwise and to a policy stance that seeks to discourage their migration. Drawing on ﬁeld research in Ghana, I argue that while rural–urban migration gives women access to an independent source of income, the emancipatory potential of migration for women is limited by the ofﬁcial stance towards rural–urban migration and informality. Nevertheless, the decision by women to migrate represents an attempt to improve their life outcomes as well as those of their families, in the face of severely constrained options for doing so. Sustainable Development Goal 5 (SDG5)—promoting gender equality and women’s empowerment—requires a different approach to women’s internal migration. Rather than seeking to constrain women’s mobility, policy and program interventions should be geared towards expanding women’s freedom to choose whether or not to migrate—by expanding the options available to women who stay at home as well as improving migration outcomes for those who migrate.


Introduction
Two significant trends in internal population movements in Africa are the re-emergence of circular migration between cities and small-holder farming areas since the 1980s [1,2], and the increasing participation of women in migration streams previously dominated by men [3][4][5][6][7][8][9][10][11][12][13].Across the continent, women account for an increasing proportion of migrant youth, with 101 female migrants under the age of 20 for every 100 male migrants [14].Although it has long been assumed that African women migrate predominantly in association with family members, a growing body of evidence suggests that more and more, they are migrating autonomously in search of employment, primarily in the urban informal economy [12,15].
Despite the fact that internal migration is quantitatively more important in developing countries than international migration, at least in terms of the sheer numbers of internal migrants worldwide, there has been relatively little research on women's migration within national borders, and with few exceptions, even less interest in understanding the dynamics of women's short-term circular migration between rural and urban areas.This is in sharp contrast to the significant interest among policy makers and researchers in the migration of women across national borders.In part this neglect is due to the fact that internal migration in general, and rural-urban migration in particular, is regarded as an economically, socially, and politically destabilizing process [16].This view leads to a misguided perception among policy makers and the general public that women who migrate from rural to urban areas are worse-off than they would have been otherwise, and to a policy stance that seeks to discourage the migration of women from rural to urban areas [17].The problem with this approach, however, is that more often than not, it only serves to increase the risks and vulnerabilities experienced by these women, which is in direct contradiction of Sustainable Development Goal 5 (SDG 5)-the promotion of women's empowerment.
This paper explores the complex links between women's empowerment and the migration of women from rural to urban areas, and argues for a reframing of policy towards women's rural-urban migration in the context of SDG 5.It is based on a study of women's migration in Ghana, where there has been an increase in the temporary migration of women moving from rural agricultural communities in the north of the country to urban centers in the south since the 1980s.This migration may be characterized as short-term circular or seasonal migration, with women leaving their origin communities for five to six months at a stretch to work in urban areas in the south, often right after the groundnut harvest is over in November or December, and returning home in April or May to help with planting activities and to gather sheanuts when they are in season, or in September to October to help with the harvest.The capital, Accra, is a major destination for most of the migrants, as are the secondary cities of Kumasi and Takoradi.
Drawing on evidence gathered from interviews and focus group discussions with women migrants in Accra and in the Savelegu-Nanton district in Northern Ghana, I argue that while poor women who migrate from rural to urban areas often experience a double burden of marginalization and exploitation, their decisions to migrate represent an attempt to improve their own life outcomes as well as those of their families, in the face of severely constrained options for doing so.For these women, the ability to return to their communities of origin, while retaining the option of coming back to the city whenever they need to do so is an important part of their efforts to exert agency in the choice of where to live and work.
Promoting the empowerment of women worldwide in developing countries entails an expansion of their ability to exercise control over major decisions that affect their lives' outcomes [18,19].This includes the ability to freely choose whether or not to migrate-a decision that requires the presence of a set of equally viable alternatives from which to choose.Sectoral and employment policies that limit women's access to economic opportunities at home together with urban policies that seek to discourage rural-urban migration and limit the livelihood opportunities of the informally employed in developing countries only serve to limit women's freedom to choose where to live and work.Rather than seeking to constrain women's mobility, policy and program interventions in the era of the SDGs and beyond should be geared towards expanding women's freedom to choose whether or not to migrate, when, and for how long-by expanding the livelihood options available to women who choose to remain at home as well as improving migration outcomes for those who move.

Literature Review
In the last few decades, the international migration of women has been the subject of intense focus by researchers and makers.A substantial body of research into the outcomes of migration for women who move across international borders has led to a growing recognition that these women face a variety of threats to their safety.This in turn has led to calls for shifts in migration policy to protect women migrants, improve their access to services in the destination countries, and facilitate their ability to make remittances, among others.
In contrast, women who migrate within national borders, and particularly those who migrate from rural to urban areas, have received little attention since the 1990s, when the Expert Group Meeting on the Feminization of Internal Migration, organized by the Population Division of the United Nations, and held in Mexico in 1991, led to a small flurry of publications acknowledging the growing participation of women in internal migration flows and calling for research to elucidate the conditions under which this phenomenon could lead to improvements in the status of women [20,21].Since then, this group of women has been conspicuously absent from international policy discussions on women's migration.For example, a background paper on gender and migration prepared for the Human Development Report on Gender and Migration [22] mentions women's internal migration only sporadically, with the most substantial discussion of internal migration focusing on internally displaced women, while the policy recommendations at the end of the document are clearly intended to improve outcomes exclusively for women who migrate across international borders [23].Indeed, with the exception of a handful of papers by researchers associated with the Migrating out of Poverty Research Consortium at the University of Sussex that emphasize the development potential of internal migration in Asia and Africa, very little has been written about the kinds of policies that would contribute to improving the outcomes of migration for women who migrate within national borders, and in particular, women who migrate from rural areas to cities in search of work [17,24,25].
This criticism is particularly relevant for Africa, where policy discussions have focused primarily on the development potential of international migration.Despite the increased participation of women in a variety of internal migration flows in sub-Saharan Africa since the 1980s [26], it is only recently that this phenomenon has even been acknowledged in high-level policy discussions relating to migration.For example, the first African Union Strategic Framework for a Policy on Migration devoted less than a page to internal migration, and made no mention of the participation of women in internal migration flows within the continent [27].In what might be taken as a sign of progress, the recently revised version of the document devotes two pages to internal migration, acknowledges the role of rural-urban migration in Africa's rapid urbanization, and calls for more research on rural-urban migration "to deepen understanding of the unique issues faced by rural-urban migrants, especially migrant women" [28].
This lack of attention to women's internal migration is perhaps unsurprising, given the uneasy place that internal migration, especially rural-urban migration, occupies in development research and praxis.In their analysis of 84 Poverty Reduction Strategy Papers (PRSPs) from 59 developing countries, Black and Sward [16] find that more than half cite the problems posed by rural-urban migration in particular for urban environments, infrastructure, employment and services.African countries are no exception: the PRSPs of the 22 African countries in their sample blamed rural-urban migration for social dislocation, urban unemployment, and the growth of urban squatter settlements.Ghana's Growth and Poverty Reduction Strategy 2006-2009 (GPRS II), for example, points to the positive development outcomes of international emigration, but barely mentions internal migration, except to state unequivocally, in a section on housing and slum upgrading, that "the growing incidence of slum development in Ghana has been the result of rural-urban migration" [29].More recently, the IOM-Ghana Strategic Migration Plan (2011-2015) describes internal migration (in a section on internal migration and irregular migration) as a "negative migration trend . . .characterized by the high rate of migration of youth from rural and smaller communities to large urban centers" which "has resulted in increased population pressure on the resources of the urban center, homelessness, health risks and diverse socioeconomic challenges for the communities of origin" [30].To combat the "problems" posed by both types of migration, the plan proposes providing "alternative livelihood and vocational skill opportunities to potential migrants and communities affected by migration" and establishing "community-based reintegration projects in high-emigration communities to encourage potential migrants to remain in their communities of origin" (ibid).While the actual content of these "alternative livelihood and vocational skill opportunities" is not made clear in this particular document, past attempts by donors, NGOs and government agencies to limit rural-urban migration have focused on promoting rural self-employment in micro-enterprises, and in the case of women, providing them with training in activities such as artisanal soap-making, hairdressing and dressmaking, and microcredit to support self-employment.By and large, these activities remain gendered and decoupled from actual employment outcomes, and thus far have had limited impacts, if any, on poverty reduction or women's empowerment [31,32].This notion that internal migration is a development "problem" helps to explain the variety of policies that have been put in place by municipal governments in many developing countries to limit internal migration, and to actively discourage rural-urban migration in particular, for example, through forced slum-clearance and mass evictions from squatter-settlements in cities like Accra, Dhaka and Harare; "closed city" policies in Jakarta that require migrants to present proof of employment and housing, and regulations in cities across Africa, Asia, and Latin America that restrict the types of informal economic activities that provide income to many rural-urban migrants [1,22].These restrictions make the lives of migrants even more difficult-not only are they unable to access basic amenities, housing, and government support at their destinations, but the variety of policies put in place to restrict their mobility makes rural-urban migrants especially vulnerable to multiple forms of harassment.This is particularly true for women who migrate from rural to urban areas, where they are typically subject to unsafe conditions and sexual harassment.Ironically, they are then treated as victims, who must be protected from the dangers of urban life.
This tendency to treat internal migration as a problem, or to ignore it completely, also helps to explain the relative dearth of female internal migration from the abundant literature on gender and migration [22].Women who migrate within national borders, in particular those who migrate from rural to urban areas, therefore, are doubly marginalized: not only are they invisible to those who advocate for the rights of women migrants, but as internal migrants, they are perceived as a problem by policy makers and non-governmental organizations.Yet, women who migrate internally for economic reasons are as vulnerable to various forms of exploitation as women who migrate internationally, and the outcomes of their migration for the women and for their households of origin are as complex as the outcomes of migration for women who migrate across national borders.

Internal Migration in Africa
Africa has a long history of both internal and international migration, but information on the characteristics of internal migration is limited.An FAO study using data obtained from World Bank household surveys and population censuses from Nigeria, Senegal, Uganda, Burkina Faso, Kenya, South Africa, Ethiopia, Malawi and Ghana suggests that internal migration remains the dominant migration flow, and that most of this migration is from rural to urban areas: over 50% of households in Africa reporting at least one internal migrant are located in rural areas, and the share of rural-urban migrants ranges from 40% in Nigeria to 55% in South Africa [33].
Women's participation in internal flows varies widely, ranging from about 12% in Burkina Faso to about 53% in Malawi (FAO, 2017) and country studies suggest that women account for an increasing proportion of internal migrants [3,14,34,35].In South Africa, for example, female migration has accounted for most of the increase in internal migratory movements of the country's African population since the 1990s [6,11].
In all the countries in the FAO study, except in Ethiopia, the most common reason given for migration is the search for better employment opportunities.Rural migrants in particular are more likely to leave their households in search of a better job.Although the FAO study suggests that African women are more likely to cite family-related reasons for their migration, in-country studies such as Lesclingand [35] and Reed, Andrzejewski [12] suggest otherwise.The FAO finding is likely to be an artifact of the study's reliance on household surveys and census data.Women who engage in short-term seasonal, temporary, or circular migration for employment tend to leave their families behind when they move.However, these short-term moves are often not captured in household surveys and population censuses, which typically define migration as long-term moves.On the other hand, women who are moving permanently are more likely to move with their families.To the extent that population censuses and household surveys capture only women's longer-term moves, women's migration will appear to be family-related.In contrast, migration studies that define migration as shorter-term movements, such as my study in Ghana, are more likely to find that women are increasingly migrating autonomously primarily for employment reasons.

Description of the Study
Ghana, like many countries in sub-Saharan Africa, has a long history of labor migration.Although emigration from Ghana has been on the rise, internal migration is the most common option for the poorest, and data from the 5th round of the Ghana Living Standards Survey show that internal migrants make up 95.8% of the country's migrants [25].The majority of these moves are from areas of relative poverty to areas of relative wealth, and 70% of internal migrants move from rural to urban areas [36].
An important migration stream in Ghana dating back to the colonial era is the migration of labor from the north of the country to the south.This migration stream was male-dominated until the 1980s, when the country saw an increase in the circulation of women between rural agricultural communities in the north and urban centers in the south [37].In 2010, women made up 52% of migrants from the Northern Region living in the capital, Accra, up from 49.8% in 2000 [38].The majority of these migrants work as market porters in urban markets in the south, carrying loads on their heads for a fee.Their clients are mainly traders and shoppers who hire them to carry goods between storage points and market centers or transport terminals, and the demand for their services is driven by the growth in vehicular congestion in the city as well as the rising importance of petty trading to urban livelihoods.This migration flow is widely considered to be an undesirable phenomenon, and existing studies have focused on the challenges faced by the women at their destination, and on the ways in which they cope with these challenges [37,39,40].This perspective has also shaped the approach of NGOs and government agencies to women's migration from northern Ghana.While NGOs have focused primarily on providing artisanal training for women, teaching them skills such as soap-making, hairdressing and dressmaking, the Ministry of Women's and Children's Affairs in 2009 launched a migrant registration program in Accra, as part of a broader program named "Operation Let's Send Them Home", the goal of which was to repatriate female migrants under the age of 16 to their origin communities in the north of the country [41].
Motivated by the paucity of research on this particular migration flow and on the implications of women's internal migration for development in sub-Saharan Africa, I spent a year in Ghana researching the reasons for and outcomes of migration for women migrants from the Northern Region of Ghana.Although the research was concluded in 2008, more recent studies confirm the relevance of the issues raised in this paper [17].My research began in Accra, where a total of 253 migrant women were selected through a snowball sampling process to respond to a survey consisting of closed and open-ended questions designed to collect detailed information about their migration decisions, their experiences in the city, and their remittance behavior.I also held in-depth interviews with 50 of these women who indicated a willingness to respond to follow-up questions about their responses to their surveys, thereby gaining more detailed information than could be obtained in the survey.These interviews were recorded, transcribed, and coded to identify central themes.For this part of my field research, I defined a migrant as any woman who had moved from the north of Ghana to live in the south for at least three months prior to the interview, or who intended to stay in the south for at least three months following the interview.This phase of my research took place in five separate locations in Accra, all chosen for their proximity to the market centers where the majority of female migrants from northern Ghana live and work.The initial sample was purposely selected to ensure diversity in terms of age, marital status, and origin community.Although the majority of women in my initial sample worked as market porters, in the later rounds I was able to interview migrant women who were employed as domestic workers, dishwashers for street food vendors, and even a waitress in a small restaurant.Although the majority of the migrants were young, single and childless, there were many married women as well as women who had at least one child (Table 1).
The second part of the field study took place in the Savelugu-Nanton district of the Northern Region of Ghana.Here, 181 households in 24 communities were randomly selected from lists of households provided by local authorities to participate in a survey that was designed to collect information on, among other things, migration, remittances, and household expenditures on education.
All the households in my sample belong to the Dagomba ethnic group, the predominant ethnic group in the Region, and the survey questions were translated into the local language, Dagbani, with the help of an interpreter.Focus group discussions were also held with different groups-community leaders, older women, younger women, older men and younger men, non-migrants and return migrants-to better understand the social context within which migration was occurring from the area.Here too, I was able to hold in-depth interviews with fifty migrant women who had returned home for the planting season and had been identified as return migrants in the household rosters.

Understanding Women's Migration from Northern Ghana
Interviews with women migrants as well as focus group discussions and interviews with community leaders and elders (both men and women) revealed that the rise in women's migration from the Northern Region of Ghana since the 1980s has been driven by the interaction of increasing livelihood insecurity with gendered social norms that give women, especially the young and single, few options for generating a livelihood at home.
The major economic activity in northern Ghana is small-holder agriculture, with 88% of households relying on crop cultivation as their chief livelihood activity, and 60% of farmers cultivating plots of less than two hectares [42,43].The sector was hit hard by the country's Economic Recovery Program of the 1980s, which abolished guaranteed minimum prices for food staples and removed subsidies to agricultural inputs, and more recently, by low and fluctuating rainfall levels [42,44].Falling yields of the major staple crops (millet, rice, maize, and sorghum), together with rising real food prices have resulted in a situation of chronic food insecurity in the region [42,43,45].According to the 2009 Comprehensive Food Security and Vulnerability Assessment of the World Food Program, 450,000 people (roughly a tenth of the population in the north of Ghana) were food-insecure in 2008, when I conducted my research [46].The situation has since worsened: in 2012, the food-insecure population had increased to 680,000, or 16% of the population [43].
A number of studies have drawn a link between migration from agricultural communities in sub-Saharan Africa and livelihood vulnerability resulting from economic crises, droughts or erratic weather patterns [47,48].While persistent out-migration from the Northern Region of Ghana certainly seems to fit this pattern, what is interesting about current migration flows between the north and south of Ghana is the increased participation of women in these flows since the 1980s.Understanding the rise of women's migration from the region requires an understanding of the ways in which economic changes have interacted with gender relations to create the conditions that support women's migration from the region.
The typical household in the Northern Region of Ghana is an extended family unit built around a closely related group of married men and their conjugal families.Food staples are collectively cultivated on the household farm, which is managed by the household head, usually the oldest male in the compound.Male household members are allocated individual parcels of farmland, on which they cultivate cash crops, and control the income generated from these private farms.Although they contribute some of this private income towards household purchases of food staples, there is no requirement that they do so, and many men will use the money from these farms to pay for purchases such as furniture and bicycles.Women's household obligations as well as their claims on household resources vary with age, seniority, and marital status.Senior married women (those with two or more children) are required to supplement the food staples with vegetables, legumes, and protein, and when stocks of the food staples run low, as has increasingly been the case, they may even become the main providers of food.Unlike men, however, they are not entitled to own or inherit land, although in land-rich households, they may be allocated a small piece of land on which they may cultivate vegetables and legumes for household consumption.To meet their obligations for household provisioning, senior married women engage in various income-generating activities such as food-processing and shea-butter extraction and are also entrusted with the sale of the crops grown collectively by the household, a service for which they expect to receive a commission [49].Although they retain full control over the income from these activities, they also face strong social obligations to use this income to meet household food needs.In contrast, junior married women, as well as unmarried women, have no access to land and no claims on household resources.However, in addition to providing unpaid labor on the household farm and within the household, they are expected to assist senior women with their income-generating activities and to contribute to the household any payments they receive in exchange for providing labor on neighboring farms.
The focus group discussions were helpful in understanding how women's responsibilities interacted with economic conditions to shape patterns of women's mobility.In response to a question about why women's migration does not appear to have started until the 1980s, I was told: In the past, men were the only ones who left during the lean season to go to farming areas in the south.At that time, women could stay home and take care of the house.Women didn't have to farm, because the men could afford to cultivate large areas of land.We women got a lot of groundnuts and rice when we helped our husbands to harvest.We could store it for food, or we could wait until the price was high before selling it for cash.Also, we could store all the sheanuts that were harvested and sell them when the price was high, so there was no need for a woman to go to the south to look for money.In those days, it was very rare to see a women migrating from this place.If this happened, everyone knew that it must have been because of very serious financial difficulties.It was degrading for a woman to go and work for money in the south.
-Head of a women's group and focus group participant, Bunglung.
Women's limited mobility before the 1980s had apparently been supported by the expansion of state-sponsored capitalist agriculture in the Northern region in the 1960s and 1970s, together with state support for agriculture more generally, both of which had generated multiplier effects that benefited women.In particular, women could count on employment on groundnut and rice farms during the planting and harvest seasons, in return for a share of the crop.Like the head of the women's group in Bunglung, many of the older women participants in the focus groups in other communities described how they had been able to earn enough before the 1980s to meet household food needs, and still have enough left over to sell, thereby meeting their other obligations and providing for their own needs as well as those of their children.
Although the impacts of the neoliberal economic policies that began in the 1980s and the subsequent decline in yields of food crops on women's livelihoods in northern Ghana have been poorly documented, there are many reasons to think that the well-being of women in the region would have suffered under the impact of the economic reforms.Following the removal of subsidies on fertilizers and other agricultural inputs in the 1990s, average farm sizes began to dwindle, as farmers found themselves unable to cultivate large tracts, and so did the yields for most crops [50].Increasingly erratic weather patterns also meant greater unpredictability in farm yields.My focus group discussions revealed that women in the region experienced these changes in a number of ways.Their in-kind payments from participating in the harvest fell, so that they now had little or nothing left over to sell after meeting household food needs.Declining yields of food crops also meant that household food supplies ran out well before the next harvest, leaving households dependent on the market to make up the short-fall.Women's obligations to make up the gap between falling stocks of food and adequate food intake led to greater diversification of women's economic activities, and women now saw migration to the south as an option for coping with increasing economic adversity: These days, farm yields are poor, and since your children have to eat, you end up using your trading capital to feed the house.We sell our sheanuts and groundnuts during the harvest season when the price is low instead of waiting for the price to rise.So now when our daughters see that we are suffering, and we can't provide their needs, they go to work in the south so that they can help us.Sometimes, women will even encourage their daughters to go in order to help out at home.
Women's economic activities play an important role in the process of livelihood diversification in rural households faced with the challenges of survival under increasingly difficult circumstances.Although social constructions of gender may constrain the activities into which women can diversify, they are not static: instead, they interact with changing economic conditions to open up new sources and opportunities for diversification that may have been previously denied to women, such as migration.Rising food insecurity in Ghana's Northern Region implies greater demands on women to supplement or replace diminished household food supplies.In the context of deteriorating agricultural livelihoods, the responsibilities of women for household provisioning, and their growing inability to meet these obligations while remaining at home, constitute an important reason for the migration of women from the region [50,51].
My interviews with women migrants in Accra confirmed that for most of them, migration to the south was driven by the need to meet their financial obligations at home or to assist senior female relatives in meeting their obligations, in the face of limited options for doing so.Salamatu (names have been changed to protect the privacy of those interviewed), a married woman in her late thirties or her early forties, had come to Accra after the petty trading business she had been operating failed.The first of her husband's two wives, she had three children and was responsible for paying their fees as well as any other expenses related to their education.When I met her, the oldest, a girl, was about to start senior secondary school and needed to buy a mattress, books, a trunk, and other items.Salamatu had already bought all these items for her with the money she made working in Accra, and she had saved over 1000 Ghana cedis (about $1000 at the time), part of which was intended to pay her daughter's fees.She explained: This year, things are even worse than before.Our farms have been destroyed (by a drought in the Northern Region that began shortly after the start of the 2007 planting season).How are we going to manage, if we don't come to Accra?I came to Accra because if I stayed at home, my children would not go to school.My husband always says he does not have money.All the money I was getting from trading was going to feed the house.I even started using my capital to buy food.So after a while there was no money left, and I decided to come to Accra.
Staying home was not an option for these migrants: as Mariama, a married woman who had returned home to Bunglung from Accra to harvest sheanuts remarked: If you stay here you will be idle, and you will eat all the food you have, and then when you run out, what will you do?Not surprisingly, seeking employment was the most important reason given by women migrants for their migration to the south, and many of the women I spoke with pointed to the lack of economic opportunity at home as a primary reason for their migration (Table 2).Women who had migrated to the south wanted to save money to meet household consumption needs, pay for children's health and education, and invest in income-generating activities (Table 3).Even when younger women migrants indicated that they wished to spend their savings on acquiring "personal" assets such as cooking pots, further questioning revealed that these were intended for use in food-processing activities such as the parboiling of rice and extraction of groundnut oil and shea-butter, the income from which would ultimately be used to meet household provisioning obligations when they acquired senior status in their households [50,52].For the majority of the women who were surveyed in Accra, circulating between their homes in the Northern Region and urban areas in the south of Ghana appeared to be part of a multi-sited livelihood strategy-a way of balancing the risks of unpredictable income in the city with the reality of limited economic opportunities at home.Almost 80% of them had been in Accra for less than a year and of those who had been in Accra for longer, 44% had returned home for periods ranging from 1-5 years.Of those who provided a reason for their most recent visit home, almost 43% reported going home to help on the family farm or to work on their own account.More importantly, almost all planned to return home, with 60% planning to stay there for more than a month and 41.2% of that number planning to help with the family farm or work on their own account upon their return.For those who had not returned, the most common reason given was a failure to meet their financial targets at the time of the interview.Clearly, these women were not migrating with plans of moving to Accra for the long-term.Instead, circulating between their rural communities and the city was an important part of their efforts to ensure that they could earn an income and meet their obligations at home.As Sanaatu, a 23-year-old migrant whom I met just a few days after she had returned home to help her mother harvest sheanuts, explained: I have been going to Accra every year for the last three years.Every year, I come back once or twice to help my mother harvest sheanuts or groundnuts.The first time I went to Accra, I didn't want to go, but it was the only way that I could make enough money to buy the things I need, and also help my mother look after my younger brothers and sisters.I come back to help my mother with the harvest because I know that I will always be able to leave something behind for her.I do this because my earnings in Accra are unpredictable.
Sometimes I have a lot of money and sometimes I don't make much money.But if I help her to harvest sheanuts or groundnuts, she will have enough left over to sell for food even if I can't send her money.
In the next section, I explore the outcomes of rural-urban migration for the women migrants I interviewed in Ghana.

Internal Migration and Development: The Role of Remittances
Although internal migration is often overlooked in discussions of poverty reduction, research shows that it can have positive impacts on household livelihoods.For example, evidence from India suggests that the sum total of internal remittances exceeds that of international remittances and that the rate of internal remittance receipts is much higher than international remittance receipts in poorer states [25].In addition, internal remittances have been shown to have a positive impact on receiving households in terms of debt payment, nutrition, education, and enterprise investment [53][54][55].In Ghana, national data from the GLSS 5 indicates that 81.8% of internal migrants sent remittances in the previous 12 months, compared with only 18.2% of international migrants, and although internal remittances were smaller in volume, they are more frequent and accounted for 53.3% of total remittances received by households [25].These remittances are especially important for the poor: for households in the lowest quartile, remittances account for a fifth of total consumption (ibid).
My own research suggests that remittances from women who migrate within national borders can have a positive impact on receiving households.Women rural-urban migrants from Ghana's Northern Region participate in female-centered networks of remittances flows, directing their remittances to female relatives in their sending households, who in turn maintained control over the use of these remittances.Thus, there was an increased likelihood that women would be the primary recipient of migrant remittances if they lived in a household from which other women had migrated to urban areas in the south of the country.The impact of these women-controlled remittances was reflected in higher education expenditure per child in households in which women were the primary recipient of remittances [50,52].

Rural-Urban Migration and Women's Empowerment
The relationship between rural-urban migration and women's empowerment is more complex.Under certain circumstances, migration can be empowering for women, and can be a way for women to escape social control or gender discrimination [1,56].In northern Ghana, the act of migrating itself represents a challenge to indigenous ideological constraints on women's mobility [57,58].In addition, the ability to send remittances can also have an impact on social norms and values, particularly those relating to gender roles and responsibilities in the migrants' communities of origin.For example, Abdul-Korah [57] shows how the ability of Dagaaba women who migrate from north-western Ghana to urban areas in the south to send money home on a regular basis has led to women taking over roles and responsibilities that were previously thought to be exclusively male, such as bearing financial responsibility for parents in their old age; this, in turn, has led to improvements in the social valuation of women at home and in the status of women in the family.
However, the extent to which the liberating potential of rural-urban migration for women themselves is tempered by multiple factors and may be realized in certain dimensions but not in others.In his discussion of the conditions under which migration can be empowering for women, Hugo [56] argues that migration is more likely to empower women if it moves them from rural to urban locations, engages them in employment outside the home in formal sector occupations, and takes place within the legal framework for an extended period.Two of these conditions-formality and legality, not only of employment, but also of housing-pose significant challenges for the realization of the empowerment potential of rural-urban migration for women in developing countries.
As in other parts of Africa, the informal economy provides the major source of employment for the majority of women rural-urban migrants in Ghana.All the women I interviewed in Accra were informally employed in the city's largest markets.The majority reported being self-employed; however, as market porters, they were effectively dependent on their clients, who paid them per load.While their earnings from such employment were higher than they would have received at home, and therefore represented some degree of economic mobility for the women, they were also highly variable.Over 90% of those interviewed reported having some savings, although these were not enough for them to move up the ladder to more stable, less dependent, and higher-earning forms of urban informal employment.Access to loans or credit from the formal banking system was non-existent: only one woman had her savings in a formal financial institution; the rest participated in rotating savings and credit associations, or kept their savings at home, with a relative or with a money-lender, and complaints of theft and loss of savings were common.
The official stance on urban informal employment represents another challenge for women who migrate to urban areas from rural areas.Despite its importance in African economies, informal employment is often perceived as "backward" and inconsistent with modern urban settings, a view that leads to efforts by municipal governments to reduce the visibility of informal employment by imposing various restrictions on informal economic activity.However, this can be disastrous for the livelihoods of the informally employed (Heintz, 2010; Skinner, 2010).The women I interviewed in Accra complained of having to pay frequent fines to the Accra Metropolitan Authority (AMA), in addition to a daily flat tax levied by the state, and of having their property and loads seized, often for reasons that were unclear to them (in 2017, the Ministry of Finance ended the practice of taxing market porters).On many occasions during my interviews, I would see women migrants running away in panic because an AMA official had been spotted in the vicinity.In some of the most egregious cases, these officials would lash out at the women with canes in an attempt to stop them from working in a particular location.Yet, these were often the same locations where there was a high demand for their services, or through which they needed to pass as they moved goods from one place to another for a client.
A third challenge for the realization of women's empowerment through rural-urban migration arises from the lack of affordable housing in many African cities, and the inability of municipal governments to plan adequately for urban growth.Finding shelter in the city was a particular problem for the majority of the women I interviewed in Accra.A third of the migrants who participated in the survey had been unable to find shelter and were compelled to spend their nights on sidewalks, shop fronts, and in bus terminals.Slightly more than half were renting a one-room wooden structure, usually in an informal settlement, which was shared with up to 20 other migrants.Only 10% were lucky enough to have access to shelter in a more permanent structure, which was often shared with others.Most were dependent on public standpipes and public toilets for water and sanitation, services for which they had to pay.As in other African cities, the threat of forced eviction was common.The tendency to blame rural-urban migration for rapid urban growth, and to see informal settlements as inconsistent with the goal of modernization has resulted in forced evictions from the settlements where many of these women lived, the most notable being the clearance of the "Old Fadama" settlement in Accra as recently as 2015 [59].
Unsurprisingly, the women migrants I surveyed in Accra were ambivalent about the emancipatory potential of migration for them.Asked whether they would recommend migration to Accra to a friend, 63% responded in the affirmative, without reservation; 12% had some reservations, and 25% said they would not.Many acknowledged the importance of having independent access to income, control over their life's decisions, and the ability to send money home to support their families or invest in some kind of income-generating activity on their return.For example, Ramatu, a married woman in her late twenties, who started going to the south to the work when she was in her teens, and continued to go even after she got married, explained: I have been lucky to get a job with a woman who buys and sells onions in bulk.I help to sell the onions, clean the shop and sweep.Initially, she was paying me 2.00 Ghana cedis a day (about $2.00 at the time of the interview) based on how many bags of onions I could sell.Now, I earn about 10.00 Ghana cedis.I can save enough money to farm when I come back.
At the same time, however, many also pointed out the contradictions inherent in the process, noting that their work as market porters was physically grueling, and complaining about their unpredictability of their income.Many noted that although they were pleased about being able to find work easily in Accra and earn some money, they also had to pay for many of the things that they had received for free at home-food, shelter, and basic services like water and sanitation.For example, Zeinabu, who was single and in her early twenties, noted how migrating to Accra had increased her sense of independence and autonomy, while at the same time exposing her to concerns that she had not had previously.
Life in Accra is better in some ways, and worse in other ways.I can work and make money here, unlike in my hometown.Also, I am no longer under my parents' control-I am free to do as I please.At home, I didn't know how to manage my finances, but now, my attitude towards saving money has changed: now I know that I have to save money to meet my goals.Before, I used to think only about getting married, but now I think that I want to work and save money first.But life here is not easy.At home, I don't have to worry about food and accommodation but here, I have to pay for everything, even to use the toilet.And the work is not easy here.I have to work long hours in the sun just to get a little money.Sometimes you may not even get enough to pay for food for the day.Also, the city officials don't allow us to stand at the bus stop waiting for customers-they chase us whenever they see us there, and make life difficult for us.And sometimes the people who hire us to carry their loads don't pay us the amount that we agreed upon.
Another migrant, Ayesha, also single and in her early twenties, pointed out how her ability to move around the city freely was curbed by her lack of familiarity with the city as well as the high cost of transportation, both of which meant that she was only able to go to places that were within walking distance; for her, she had enjoyed much greater freedom of movement at home than she had in the city: Life in Accra is different.Although I can work and earn money here, I am not free to go wherever I want.I don't have anywhere to go, because I don't know a lot of places in Accra, and I have to work all the time.The only places I know are the market where I work, and the place where I sleep.Even if I wanted to go somewhere, I don't have money to pay for a trotro (a mini-bus that is a popular form of transport in the city).
Clearly, for all these women, periodic migration to the south was an important element in a multi-sited livelihood strategy, and the outcome of their efforts to ensure a sustained livelihood, in the face of economic and social constraints on their access to the resources for doing so at home.In making the decision to migrate, women were exercising their agency, even as they were very much aware of the risks and challenges that they could be exposed to in the migration process.Given that the promotion of gender equality and women's empowerment is Goal 5 of the UN's Sustainable Development Goals, what is the appropriate role of development policy towards the increased participation of women in rural-urban migration flows in Africa?

Towards a Sustainable Approach to Women's Rural-Urban Migration
Current policy discussions on migration, which have been primarily concerned with cross-border movements, have focused either on improving outcomes for migrants at their destinations, or on encouraging development at home by supporting rural development projects.Bakewell [60] has argued forcefully against the latter, noting that the desire to reduce rural-urban migration by improving conditions in rural areas constitutes a primary reason for investments in rural development programs by African governments, international development NGOs and donors.These programs, he argues, are grounded in an approach that assumes that development is about enabling people to stay at home without taking into account whether or not they actually wish to do so.He writes: " . . .development action to sustain some rural areas, or even whole countries, may be attempting to create artificial incentives to keep people in their place.In some cases, it could be asked if investing such aid is wasteful when migration may be a more attractive and sustainable option for those people who have the opportunity to take it."(Bakewell, 2008(Bakewell, , p. 1353) ) For Bakewell, the focus should be on improving outcomes for migrants at their destinations.In the context of rural-urban migration, this would involve improving outcomes for migrants living in urban areas, an approach that has also been advocated by researchers affiliated with DFID's Migrating out of Poverty Research Consortium [1,24,25].
On the other hand, Zoomers and Adepoju [61], writing about the limits of international migration for sustainable development, note that despite the growing attention to international migration, the majority of the population in Africa is geographically stable, and argue for more explicit attention to be paid to improving local employment opportunities for those who stay, especially the youth.Presumably, this would involve expanding employment opportunities in both rural and urban areas in countries of origin.However, if their argument is extended to rural-urban migration, it would require the expansion of employment opportunities primarily in rural areas, precisely the approach that Bakewell criticizes.Are these two approaches to "managing migration" mutually exclusive?A focus on the needs of women migrants who straddle both rural and urban locations in Africa can perhaps provide some clarity for the direction of policy.As my research in Ghana suggests, Bakewell's argument that migration may be "more attractive and sustainable" for those who choose to migrate overlooks the possibility that migration is not always freely chosen: people may migrate because they feel they have no other options.The migrants I interviewed were attracted by the employment opportunities provided by urban locations, but nevertheless did not plan to move permanently, choosing instead to maintain a significant presence in their rural communities.Given the conditions under which their migration currently occurs, this is perhaps not surprising.Migration represents an opportunity for women to improve their life outcomes in the face of severely constrained options for doing so, but many of the women I spoke to made it clear that they would not have chosen to migrate if they could have gained access to the economic opportunities that they sought while remaining at home.For this group of women, policies that reduce their access to economic opportunity at home only serve to further limit their options for sustaining a livelihood without migrating.On the other hand, it is also quite possible that there are many women migrants who would have been happy to remain longer in urban areas had conditions in these urban locations been better.For this second group of women, as well as the first, policies that restrict informal economic activity and the growth of informal settlements in the city worsen the risks they face, leading to greater insecurity and higher risk of sexual harassment, exploitation, and abuse as a result.For both groups of women, ultimately, the goal of policy should be to ensure that migration is freely chosen-in other words, that it is chosen from among a set of equally viable alternatives and that women who choose to migrate or to stay at home could have chosen differently, had they wished to do so.
What kinds of policy and program interventions-with regard to agriculture, employment creation, urban planning and the informal economy-can help to expand the freedom of poor rural women to choose whether or not to migrate?What kinds of policies can help to improve outcomes for those women who choose to migrate, as well as those for those who choose to remain at home?These are questions that are not easily answered, but it is possible to identify some general directions for policy by understanding the ways in which current neoliberal economic policies have contributed to the present state of affairs.
The rising participation of African women in migration flows previously dominated by men, as well as their growing participation in labor circulation between rural and urban areas rather than in permanent migration needs to be understood in the context of the failure of decades of neoliberal economic policy in Africa to generate adequate employment opportunities in rural and urban areas.Market liberalization, the withdrawal of state support from agriculture and increasing competitive pressures from global markets, has resulted in the growing inability of rural households to depend on farming alone for their livelihoods and the growing proletarianization of smallholder farmers and their families [2,31].In this context, migration to urban areas and employment in non-agricultural economic activities provides access to higher earnings than women will be able to receive were they to remain at home.Yet, permanent migration does not offer sustained upward mobility for many of these women.In the post-liberalization context of deindustrialization, women face bleak prospects for good employment in towns and cities, and so it becomes even more important to maintain strong attachments to home, thereby maintaining economic activities and residence in multiple locations.This lack of security in urban destinations is further heightened by the urban management approach of African municipal governments committed to creating "world class cities" in which informal employment and housing are seen as backward and inconsistent with modernity.
In the absence of adequate demand for labor in rural areas combined with inadequate employment opportunities in urban areas, the phenomenon of women floating between rural areas and urban areas in coping migration patterns can only persist.A sustainable and empowering approach to this development dilemma must involve policies that expand women's employment opportunities in their areas of origin as well as at their destination, together with a shift in perceptions regarding the role of rural-urban migration and the informal economy in African development and a reorientation of economic policy and municipal planning towards improving working and living conditions in the urban informal economy [62].Expanding women's employment opportunities in the areas of origin will require agricultural policies and rural development policies that seek to create rural employment for those, including women, who are currently unable to generate a sustainable livelihood from small-scale farming alone.Improving working and living conditions in the urban informal economy will require a policy approach towards informal employment, social protection, and urban employment generation that goes beyond microloans, as well as more progressive solutions for housing the poor in urban areas.In Ghana, some steps are being taken in the right direction: a health policy dialogue facilitated between the National Health Insurance Authority, the Ministry of Health, and market porters resulted in over 1000 women migrants from northern Ghana gaining access to health care services through the Ghana National Health Insurance Scheme [63].While this is a start, much more remains to be done.

Conclusions
Despite the long-standing tendency to view internal migration, and rural-urban migration in particular, as a destabilizing process, it is important to remember that migration itself is not inherently good or bad.What matters, from a human development perspective, is whether it is freely chosen, or whether people feel compelled to migrate into desperate conditions because they see no other alternative.Gender norms assign significant responsibilities to women, while restricting their access to the resources they need to meet these responsibilities.In the context of declining agricultural livelihoods, migration to urban areas holds the potential to give women access to the resources they need to meet these obligations and ultimately challenge the inequalities inherent in these social norms.However, the conditions under which women live and work in their urban destinations prevents this potential from being fully realized, and mostly serves to emphasize how limited women's choices are.
The goal of ensuring that when women make the decision to move or stay, they are indeed choosing from among an expanded set of viable alternatives is one that is entirely in line with the idea of human development as an expansion in the freedom of people to live the lives that they value.By enhancing the ability of women and girls to make strategic life choices, it is also completely in consonance with Sustainable Development Goal 5, to empower all women and girls.If African governments hope to even come close to realizing this goal by 2030, African governments as well as their development partners will have to contend with the issues raised by the feminization of internal migration in the continent.

Table 2 .
Primary reason for migration (respondents were asked to identify their primary reason for moving to Accra).

Table 3 .
Financial goals of women migrants (respondents were able to give multiple responses to this question).