SOAS University of London

CDPR Development Viewpoint

Breaking the Multiple Constraints on Poor Women's Livelihoods

by Naila Kabeer, Professor of Development Studies, SOAS; Karishma Huda, Research Manager, and Sandeep Kaur, Communications and Monitoring Officer, BRAC Development Institute; and Nicolina Lamhauge, Policy Analyst, OECD Environment Directorate


Development Viewpoint Number 73, May 2012

This Development Viewpoint highlights the provocative findings and lessons in CDPR’s new Discussion Paper #28/12, 'Productive Safety Nets for Women in Extreme Poverty: Lessons from Pilot Projects in India and Pakistan'.

The Discussion Paper assesses the success of two pilot projects—one in West Bengal, India and the other in Sindh, Pakistan—in ‘graduating’ poor women out of extreme poverty. The projects were based on the successful experience of BRAC, the well-known NGO in Bangladesh, in overcoming the multiple constraints blocking women’s empowerment.

These projects recognized that extreme poverty is not based solely on a lack of material resources, such as income and assets. It is also invariably due to a lack of human resources (such as education and skills) and social resources (such as supportive kinship relations and community solidarity).

In South Asia, social inequalities based on caste, religion and ethnicity worsen economic deprivation. Thus, people from lower castes, tribal groups and minority religions are more likely to be extremely poor.

Gender discrimination cuts across these various dimensions of social inequality, consigning women from minority groups to an even more vulnerable position. Thus, they find it exceedingly difficult to escape conditions of extreme deprivation. And many projects designed to help them correspondingly fail.

The BRAC Approach

In response, BRAC has pioneered a project approach (‘Targeting the Ultra-Poor’ or TUP) that recognizes that the extreme poverty of women results from a number of intersecting constraints. Thus, it designs interventions to address them simultaneously.

These constraints include: a lack of viable livelihood options; an inability to move to where other options are available; a dearth of skills and knowledge; difficulties, under the stress of daily survival, in accumulating investment funds; a lack of self-confidence; and a reluctance to take risks.

In tackling such multiple constraints, BRAC devised an approach that included seven major elements. First, it used ‘mixed method’ surveying that combined various intersecting ways to successfully identify the extreme poor. Second, it focused on enterprise development since poor women are unlikely to ever generate enough income on their own to invest in a micro-enterprise. This aspect usually took the form of some kind of asset transfer to get an enterprise off the ground.

Third, BRAC insisted on intensive interaction and training so that poor women could boost their self-confidence and effectively manage their enterprises. Fourth, women were provided with an interim cash stipend until their enterprise could begin generating a sustainable income.

Photo: G.M.B Akash/ Panos
Photo: G.M.B Akash/ Panos


Fifth, women had to enroll in a weekly savings programme in order to accumulate a small fund to deal with inevitable shocks and stresses. Sixth, the programme offered health support based on the recognition that the costs of ill health are often a major drain on poor women’s meager resources. And, lastly, BRAC sought to mobilize elite support for its projects, usually by setting up a Village Assistance Committee composed, for example, of government officials, businesspeople and teachers.

Outcomes for the Two Pilots

BRAC’s success led to the piloting of a number of other similarly designed projects. Among these are the two that the CDPR Discussion Paper examines. These initiatives were supported by CGAP and Ford Foundation.

The project in West Bengal was managed by Trickle Up, an international NGO, in partnership with the Human Development Centre, a local NGO. The productive assets that it provided poor women were goats and sheep. In Sindh, Pakistan, the Orangi Charitable Trust (OCT) concentrated on providing goats, hens and cash for purchasing material for basket-making (the traditional activity for women in this area).

Because the Sindh project did not adopt BRAC’s mixed-method targeting, it had difficulties in identifying the extreme poor in the project villages. Also, it did not carry out the accompanying research needed to identify the viability of their livelihood options. For example, the goats that it provided died because they were not suited to the local conditions and veterinary services were inadequate.

As a result of such factors, pre-existing inequalities in the Sindh project villages tended to deepen. The better endowed households in the wealthier villages progressed the most rapidly. Pervasive patriarchal constraints also held back many of the women.

Social networks were strong within these villages, which had been settled generations ago. Frequent cross-cousin marriages meant that most families were related to each other. And there were few cases of divorce or desertion.

However, such strong kinship and family ties went hand-in-hand with a deeply entrenched preference for male breadwinners. Hence, women had difficulties in exercising the independence necessary to benefit from the opportunities offered by the OCT project. The most adverse impact was felt by families without a regular male breadwinner.

More Positive Results in West Bengal

The results in West Bengal were more positive. Following more closely the BRAC targeting approach, the programme succeeded in including a much larger proportion of the extreme poor. However, like the Sindh pilot, it failed to adequately take account of local conditions. Hence, some of the livestock that it distributed died.

Nevertheless, households generally made much greater progress in the West Bengal context, including those from the least advantaged tribal groups. There are several explanations for this contrast.

Poorer women, including many Muslim women, enjoyed much greater mobility in the public domain than women in Sindh.  Indeed, women from tribal and lower caste groups had a long tradition of working for their living. Thus, not only were they better able to respond to project opportunities but also they were able to do so even in the absence of—or in the face of resistance by—male household members.

An additional advantage of the programme in West Bengal is that many of its staff had previously worked with an Indian NGO that had lengthy experience in supporting Self-Help Groups for women in extreme poverty.

Though not included in BRAC’s original project design, this approach was introduced into the West Bengal pilot and proved to be a major strength. This group structure allowed women to save on a more regular and disciplined basis and offered a venue where they could discuss their problems and seek support.

Finally, the West Bengal pilot operated in a context in which the state was far more active than in Sindh. Also, the state’s efforts worked in tandem with—rather than in isolation from—the efforts of NGOs.

Though neither pilot carried out a baseline survey in order to rigorously judge its success, the West Bengal project gathered enough relevant information to amply support its decision to expand its outreach to other villages, and even to other states in India. However, based on its evident lack of success, the Sindh project chose to end its activities.

Concluding Remarks

The more positive outcomes for the West Bengal pilot are likely due, in part, to its staff’s experience. It was able to draw on their long track record of working with the extreme poor and understanding their livelihood difficulties.

Thus, it grasped the importance of combining survey and participatory methods in identifying the extreme poor and understood the value of the group structure in providing critically needed support to poor women.

By contrast, the Orangi Charitable Trust in Sindh might have been disadvantaged by a predominantly microfinance background. Thus, it was less accustomed to grasping the importance of the social and cultural dimensions of poverty.

The OCT project also had to work in a context in which the state was largely absent. In West Bengal, in contrast, the state was very active. The downside of such a context was the importance that had to be attached to strengthening political connections.

But such links proved invaluable in providing access to state benefits, health care and infrastructure. This advantage tends to support, for example, BRAC’s original emphasis on ‘mobilizing elite support’.

The Sindh project also operated in a social context in which kinship structures were far more stable and patriarchal constraints correspondingly stronger. Because families in the West Bengal pilot, by contrast, came from the poorest segments of the population, they had undergone considerable migration and resettlement. Thus, patriarchal constraints were weaker.

The men in West Bengal did display greater ‘irresponsibility’ toward their familial obligations (e.g., drinking, remarrying, beating their wives or deserting them) than men in Sindh. As a result, women had learned—perhaps paradoxically—to be much less reliant on men’s incomes for their own survival.

The original BRAC project design assumes, implicitly, that social conditions are flexible enough to allow women to exercise a significant degree of independence. When such conditions are clearly not present, it becomes crucially important to implement pro-active measures, such as the innovation of the Bengali self-help groups, in order to ensure that poor women are empowered to take some control of their lives.

Otherwise, projects that intend to ‘graduate’ women out of extreme poverty will find that the multiple constraints impeding women’s empowerment will inevitably reassert their regressive influence.