Addressing invisibility, inferiority, and powerlessness to achieve gains in maternal health for ultra-poor women.

Despite a continued stated commitment to social justice and equity—the guiding spirit of the Millennium Declaration in 2000—concerns have arisen that this focus has often been diluted in eff orts to translate the Millennium Development Goals (MDGs) into actions. Nowhere is this more apparent than in relation to MDG 5. Analyses of national survey data and local programme assessments show that policy directives and interventions often fail to reach the poorest women within local populations. In Pakistan, for example, a ten-district intervention aimed at upgrading health facilities while simultaneously increasing demand through behavioural change resulted in a rise in institutional deliveries in the highest wealth quintile (from 62% to 74%), but no change in the poorest (remaining at roughly 18%). Meanwhile, a crossnational analysis identifi ed large inequalities between wealth groups in access to skilled birth attendance and antenatal care uptake in many countries. Thus, despite repeated assertions that reducing inequity is a top priority, international organisations, donor agencies, and country-level policy makers continue to fail to address the obstacles to care faced by the poorest women. Furthermore, despite reductions in maternal mortality in several countries, comprehensive reviews of progress towards MDG 5 conclude that the pace is slow and that no overall evidence of acceleration exists, suggesting that unless concerted action is taken to understand and address the needs of the most vulnerable women, the aggregate goal will remain elusive in many countries. The structures and processes restricting poor people’s access to resources, including health care, are complex and intractable. The notion of structural violence, usually attributed to Johan Galtung in the 1960s, drew attention to the way in which embedded social struc tures supported by normalising ideologies system atically oppress and exploit particular groups within societies resulting in poor health and short life-spans within these populations. Naila Kabeer has used the notion of intersecting inequalities to describe how several and interlocking processes—poverty, ethnic origin, caste, and religion— converge with sex to place some women in very marginalised positions. Our research in rural Punjab, Pakistan, has documented an entrenched caste-based social hierarchy in which the lowest caste, the Kammis, experience chronic inter generational poverty, social stigma, and poor health and nutrition. We suggest that such women can usefully be identifi ed as “ultra-poor” in south Asia since the con vergence of social stigma and economic poverty places them in a very marginalised position. Importantly, we noted low uptake of maternal health-care services and very high rates of maternal mortality in these ultra-poor, socially marginalised women. Drawing on our empirical work, we emphasise key characteristics of the lives of socially marginalised women that directly aff ect their access to health services and resources, which, although specifi c, are also evident in diverse contexts around the world. Ultra-poor, socially excluded women are often invisible to both more advantaged local people and to service providers. While living in our Punjabi village, the identifi cation of all the low caste women took us 6 months. Our requests to local people to provide help to locate the poor did not identify any Kammi families. Higher-caste villagers chided us for our interest in Kammi women, and Kammi women themselves conformed to the normative expectation of remaining at the margins of society: silent and unseen. So invisible were these women that even a private non-governmental organisation working for gender empowerment and human rights in the village had overlooked them in its projects. Kammi women working as bonded labourers in nearby brick kilns were wholly excluded from statefunded door-to-door outreach maternal and child health services because the demarcation of Lady Health Worker catchment areas does not include these areas. Economic hierarchies are often closely aligned with social hierarchies with poor women occupying stigmatising social positions. In our fi eld site, the inferior social status of the Kammis was symbolised in their caste label, “kam”, which translates as less and signals the widely held belief that they have a low level of virtue and moral character. One common result of social inferiority is that such women are poorly treated by health-care providers, which acts as a great deterrent to the seeking of care. Kammi women reported that government health staff ignored and abused them, calling state hospitals “butchers”, and preferring to “die at home” rather than seek such care. And, although our fi eld site had geographically available and reasonably well functioning public-sector and private-sector maternal health services, less than a quarter of Kammi women had ever sought care from these facilities compared with near universal use by the upper castes. Socially excluded groups are often denied participation in the decision-making processes of their communities and societies and their formal entitlements to state resources can be usurped by more powerful groups. We noted that despite formal criteria for the dispersal of funds from the Benazir Income Support Programme, in practice resource allocation was decided by the local Lancet 2014; 383: 1095–97